Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Preparation and characterization of paclitaxel-loaded poly(d,l-lactide-co-glycolide) microspheres for… Guan, Dechi 1999

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_1999-0502.pdf [ 15.88MB ]
Metadata
JSON: 831-1.0089168.json
JSON-LD: 831-1.0089168-ld.json
RDF/XML (Pretty): 831-1.0089168-rdf.xml
RDF/JSON: 831-1.0089168-rdf.json
Turtle: 831-1.0089168-turtle.txt
N-Triples: 831-1.0089168-rdf-ntriples.txt
Original Record: 831-1.0089168-source.json
Full Text
831-1.0089168-fulltext.txt
Citation
831-1.0089168.ris

Full Text

PREPARATION AND CHARACTERIZATION OF P A C L I T A X E L - L O A D E D POLY(Z),£-LACTIDE-CO-GLYCOLIDE) M I C R O S P H E R E S F O R I N T R A ARTICULAR INJECTION  by DECHIGUAN  B.Sc. (Chemistry), M.Sc. (Polymer Sciences), Zhongshan University, 1984, 1987  A T H E S I S S U B M I T T E D IN P A R T I A L F U L F I L M E N T O F T H E R E Q U I R E M E N T F O R T H E D E G R E E OF M A S T E R OF SCIENCE  in  T H E F A C U L T Y OF G R A D U A T E STUDIES  (Faculty of Pharmaceutical Sciences)  W e accept this thesis as confirming to the required standard  T H E UNIVERSITY OF BRITISH C O L U M B I A July 1999 © D e c h i Guan, 1999  In p r e s e n t i n g t h i s t h e s i s i n p a r t i a l f u l f i l m e n t of the requirements f o r an advanced degree at the U n i v e r s i t y of B r i t i s h Columbia, I agree t h a t the L i b r a r y s h a l l make i t f r e e l y a v a i l a b l e f o r r e f e r e n c e and study. I f u r t h e r agree t h a t p e r m i s s i o n f o r e x t e n s i v e c o p y i n g of t h i s t h e s i s f o r s c h o l a r l y purposes may be g r a n t e d by the head of my department or by h i s or her r e p r e s e n t a t i v e s . I t i s u n d e r s t o o d t h a t c o p y i n g or p u b l i c a t i o n of t h i s t h e s i s f o r f i n a n c i a l g a i n s h a l l not be a l l o w e d without my w r i t t e n p e r m i s s i o n .  Department of QNA.CXjr>J<>C^O<aJU^Cc<x\ The U n i v e r s i t y of B r i t i s h Vancouver, Canada  Date  JSgffV >~T ,  Columbia  ScA-^AAOeS>  11  ABSTRACT Paclitaxel has been shown to cause significant regression of existing rheumatoid arthritis and to prevent the induction of collagen-induced arthritis (CIA)  in animal  models. Paclitaxel suppresses arthritis because rapidly proliferating inflammatory pannus cells in the joint are susceptible to the phase-specific cytotoxic effects of paclitaxel. Intra-articular therapy using anti-inflammatory  steroids is used for patients in  whom rheumatoid arthritis manifests itself in only a limited number of joints. The objective of the research was to prepare and characterize paclitaxel-loaded microspheres using lactide:glycolide ( L A : G A ) polymers, which might potentially be suitable for the intra-articular delivery of paclitaxel in arthritis. Paclitaxel-loaded prepared using the  poly(J,/-lactide-co-glycolide)  solvent  evaporation  method.  (PLG)  microspheres  P L G polymers having  were  different  compositions of lactide and glycolide as well as having different molecular weights with the same lactide and glycolide composition were chosen to study the influences of these factors on the paclitaxel release rate. The effects of paclitaxel loading in the polymer matrix and the sizes of the microspheres on the paclitaxel in vitro release behavior were also assessed. Paclitaxel was loaded into P L G microspheres with encapsulation efficiencies of over 90% due to the hydrophobicity of the drug. Differential scanning calorimetry (DSC) thermograms indicated that the glass transition temperatures increased with an increase in paclitaxel loading in the P L G matrices, which was believed to be due to an interaction involving the formation of hydrogen bonds between paclitaxel and P L G polymers. X-ray diffraction data showed only the presence of an amorphous matrix, with no evidence by  Ill either X-ray diffraction or D S C , of crystalline paclitaxel present in the microspheres matrix. Degradation studies of both control and paclitaxel-loaded microspheres in phosphate buffered saline (PBS) containing albumin at 3 7 ° C showed that the molecular weights of P L G microspheres with a 50:50 lactide:glycolide composition decreased rapidly with time. The molecular weights of P L G microspheres with higher lactide content (> 50 mole% o f lactide) did not decrease significantly until after 3 weeks of incubation in PBS-albumin. The release profiles of paclitaxel from all P L G microsphere formulations showed a burst phase of release, followed by a phase of relatively steady release. The burst phase was caused by rapid release of paclitaxel from the superficial surface layers of the microspheres. The release rates of paclitaxel  from PLG50:50 microspheres were  influenced by paclitaxel loading and molecular weights of the PLG50:50 polymers. Increased loading and decreased molecular weight led to faster paclitaxel release rates. P L G microspheres prepared from polymers with L A : G A ratios of 85:15, 75:25 and 65:35 showed that the L A : G A compositions had minimal effect on paclitaxel release rates. The two size ranges of microspheres showed minimal effects on the rates of paclitaxel releases from the microspheres.  IV  TABLE OF CONTENTS ABSTRACT  ii  TABLE OF CONTENTS  iv  LIST OF FIGURES  xi  LIST OF T A B L E S  xix  LIST OF ABREVIATIONS  xxi  ACKNOWLEDGMENT  xxiii  1.  INTRODUCTION  1  2.  BACKGROUND  3  PACLITAXEL  3  2.1.  2.1.1.  Source  3  2.1.2. Chemistry of paclitaxel  3  2.1.3. Crystal structure  5  2.1.4. Solubility and stability  5  2.2.  P A C L I T A X E L IN RHEUMATOID ARTHRITIS  6  2.2.1. Rheumatoid arthritis (RA)  6  2.2.2. Mechanism of action ofpaclitaxel  7  2.2.3. Preclinical effectiveness of paclitaxel in rheumatoid arthritis 9 2.3.  INTRA-ARTICULAR T H E R A P Y  10  V  2.3.1.  Side effects of paclitaxel systemic therapy  10  2.3.2.  Intra-articular therapy  10  2.3.3.  Problems associated with intra-articular  2.3.4.  Controlled release intra-articular  therapy  11  drug delivery systems 11  2.3.5. 2.4.  Controlled release microspheres for arthritis  P O L Y M E R CHEMISTRY  2.4.1.  Structure  12 13  13  2.4.1.1.  Configuration and conformation  13  2.4.1.2.  Morphology and models of polymer structures  14  2.4.2.  Molecular weight  18  2.4.3.  Biodegradable polymers  20  2.4.4.  Glass transition  20  2.5.  2.4.4.1.  Glass transition temperature (Tg)  20  2.4.4.2.  Factors influencing glass transition temperature  21  POLY(A^-LACTIDE-CO-GLYCOLIDE) ( P L G )  23  2.5.1.  Structure and synthesis  23  2.5.2.  Drug delivery systems (DDS) based on PLG  25  2.5.2.1.  Microspheres  25  2.5.2.2.  Biocompatibility  25  2.5.2.3.  Mechanism of in vitro degradation and erosion  26  2.5.2.4.  Factors affecting PLG degradation  27  2.5.2.5.  Mechanism of drug release from PLG matrix  28  vi  2.5.2.6.  Factors affecting drug release from PLG matrix  28  2.5.3. Enzymatic degradation  29  2.5.4. Effects of sterilization on microspheres  29  2.5.5. Paclitaxel-loaded PLG microspheres  30  2.6.  3. 3.1.  OBJECTIVES  31  EXPERIMENTAL  32  MATERIALS & SUPPLIES  32  3.1.1. Lactide:glycolide polymers  32  3.1.2. Paclitaxel  32  3.1.3. Chemicals & solvents  32  3.1.4. Buffers, mobile phases and PVA solution  33  3.1.5. GPC analysis  33  3.1.6. Glassware  34  3.2.  EQUIPMENT  34  3.2.1. Apparatus for microsphere manufacture  34  3.2.2. High performance liquid chromatography (HPLC)  34  3.2.3. Gel permeation chromatography (GPC)  34  3.2.4. Differential scanning calorimetry (DSC)  35  3.2.5. Scanning electron microscopy (SEM)  35  3.2.6. Fourier transform infra-red spectroscopy (FTIR)  35  3.2.7.  35  Particle size analysis  3.2.8. Polymer compositions  36  3.2.9. X-ray powder diffraction  36  Vll 3.2.10. Centrifuges  36  3.2.11. Incubator and ovens  36  3.2.12. Other equipment  36  .3.  METHODS  38  3.3.1.  Preparation of microspheres  35  3.3.2.  Characterization of microspheres  39  3.3.2.1.  Nuclear magnetic resonance  39  3.3.2.2.  Measurement of molecular weight by gel permeation  chromatography 39  3.3.2.3.  Determination of sizes and morphology of microspheres  40  A. Optical microscope  40  B. Laser scattering  40  C. Scanning electron microscope  40  3.3.2.4.  X-ray diffraction  41  3.3.2.5.  Thermal analysis  41  3.3.2.6.  Fourier transform infra-red spectroscopy (FTIR)  41  3.3.3.  Total content of paclitaxel in microspheres  3.3.3.1.  Determination of total content ofpaclitaxel  3.3.3.2.  Recovery ofpaclitaxel from microspheres  3.3.4.  42 in microspheres  43  Paclitaxel release studies  43  3.3.4.1.  Paclitaxel standard curves used for HPLC analysis  3.3.4.2.  Determination of chemical stability ofpaclitaxel albumin, distilled water and methanol  42  43  in acetonitrile, PBS44  Vlll  3.3.4.3.  Determination of in vitro release profiles ofpaclitaxel from microspheres  44  3.3.5. Microsphere degradation studies  45  3.3.6. Co-60 irradiation of PLG microspheres and paclitaxel  46  3.3.7. Statistical treatment of data  46  RESULTS  48  4.1.  P A C L I T A X E L A N D 7-EPI-TAXOL  48  4.2.  P H Y S I C A L CHARACTERIZATION OF P L G POLYMERS  48  4.2.1.  The molar compositions of PLG copolymers  4.2.2. Standard GPC curve and molecular weights of polymers  48 49  4.3.  SIZE DISTRIBUTION OF T H E MICROSPHERES  55  4.4.  S U R F A C E MORPHOLOGIES OF MICROSPHERES  55  4.5.  C H A R A C T E R I Z A T I O N OF MICROSPHERES  64  4.5.1. X-ray diffraction patterns of the microspheres  64  4.5.2. Glass transition temperatures of control and paclitaxel-loaded microspheres 4.5.3. Fourier transform infra-red spectroscopy (FTIR)  67 67  4.5.4. Total content and encapsulation efficiency of paclitaxel in microspheres  75 4.5.5. Stability of paclitaxel in acetonitrile, PBS-albumin, distilled water and methanol 4.5.6. Release studies of paclitaxel-loaded microspheres 4.5.6.1.  Statistical treatment of release study data  75 79 79  ix  4.6.  4.5.6.2. Paclitaxel release from PLG microspheres  79  IN VITRO DEGRADATION OF MICROSPHERES IN P B S - A L B U M I N B U F F E R  89  4.6.1. Molecular weights of degraded PLG microspheres determined by GPC 89 4.6.2. Surface morphology by SEM. 4.6.3. Glass transition temperatures of the degraded PLG microspheres 4.7.  100 105  T H E E F F E C T OF Y-IRRADIATION O N P A C L I T A X E L , C O N T R O L A N D P A C L I T A X E L L O A D E D MICROSPHERES  107  DISCUSSION  110  5.1.  COMPOSITION OF T H E P L G POLYMERS  110  5.2.  MICROSPHERE FABRICATION  110  5.  5.2.1. Size and surface morphology of the microspheres  110  5.2.2. X-ray diffraction patterns of control and paclitaxel loaded microspheres 112 5.2.3. Paclitaxel total content and encapsulation efficiency  113  5.2.4. Glass transition temperatures of control and paclitaxel-loaded microspheres  113  5.2.5. Investigation ofpolymer-paclitaxel interaction by Fourier transform infrared spectroscopy (FTIR) 5.2.6. In vitro degradation of microspheres 5.2.6.1.  Molecular weights and morphological changes  5.2.6.2.  The effect of degradation on glass transition temperatures  114 116 ....116 119  X  5.2.7. In vitro release ofpaclitaxel from paclitaxel-loaded microspheres 119 5.2.8. Effects of gamma irradiation on paclitaxel and PLG polymers 5.3.  123  INTRA-ARTICULAR MICROSPHERE F O R M U L A T I O N A N D F U T U R E WORK  124  6.  S U M M A R Y AND CONCLUSIONS  125  7.  REFERENCES  127  XI  LIST O F FIGURES Figure 1  Chemical structure of paclitaxel (numbers assigned to the carbons in the structure according to T U P A C nomenclature) (The Merck Index, 1996) 4  Figure 2  Illustration of normal and inflammatory joints (adapted from Harris 1997). 8  Figure 3  Fringed-micelle, folded chain and switchboard models (adapted from Rabek, 1980)  Figure 4  15  Single-phase random-coil and folded-chain fringed micellar grain models (G: grain; IG: intergrain; G B : grain boundary and O D : ordered domain) (adapted from Sperling, 1986)  Figure 5  17  Molecular weight distribution of a polymer (Rabek, 1980) 19  Figure 6  D S C thermogram for a polymer illustrating the different, possible thermal events (Rabek, 1980)  Figure 7  22  Synthetic pathway for P L G polymers and chemical structure of stannous 2ethyl-hexanoate catalyst (Zhang et al., 1994) 24  Figure 8  Illustration of microsphere fabrication (A: D y n a - M i x overhead stirrer with a speed controller and B: 4-blades overhead impeller) 37  Xll Figure 9  S E M micrographs of A : paclitaxel crystal (magnification: 5.0kx at l O k V , 25mA) and B : 7-epitaxol crystal (magnification: 5.0kx at l O k V , 25mA) 51  Figure 10  Representative H ' - N M R (200MHZ) spectra o f P L G polymers [A: a P L G polymer with a L A : G A composition of 100:0 and I.V.= 0.60, B: a P L G polymer with a L A : G A composition of 75:25 and I.V.= 0.55, and C : a P L G polymer with a L A : G A composition of 50:50 and I.V.= 0.78. Solvent used: Chloroform-D (99.8%+)]  Figure 11  52  G P C elution profiles of polystyrene standards with molecular weights ranging from 2,000 to 100,000 g/mole at 4 0 ° C (Solvent used: chloroform. Solvent flow rate: 1.0 mL/min. Injection volume: 20u.L) 53  Figure 12  G P C standard curve for polystyrene standards with molecular weights of 4, 9, 17.5, 30, 50, 100 and 170 kg/mole on a PLgel column with a nominal pore size of 10 A at 4 0 ° C (Solvent used: chloroform. Solvent flow rate: 4  1.0 mL/min. Injection volume: 20 u.L) Figure 13  54  Representative size distributions of A : small and B : large sizes of microspheres determined by the laser scattering particle size analyzer. The size distributions of the microspheres were the averages of the two 60seconds scans. Microspheres were resuspended in polysorbate 80 solution. 56  Figure 14  S E M micrographs of 20% paclitaxel-loaded P L G microspheres [A: PLG100:0, 20-lOOum (lkx at 5kV, 25mA); B: PLG85:15, l-20um (lOkx  Xlll at 6kV, 25mA); C : PLG75:25, l-20um (4kx at 6kV, 25mA); and D: PLG65:35, l-20um (lkx at 5kV, 25mA)] Figure 15  57  S E M micrographs of 10% paclitaxel-loaded P L G microspheres [A: PLG100:0, 20-100pm (2kx at 5kV, 25mA); B: PLG85:15, l-20um (lOkx at 6kV, 25mA); C : PLG75:25, l-20um (600x at 5kV, 25mA); and D: PLG65:35, l-20um (lOkx at 6kV, 25mA)]  Figure 16  58  S E M micrographs of control P L G microspheres [A: PLG100:0, 20-lOOu.m (1.5kx at 5kV, 25mA); B: PLG85:15, l-20um (6kx at 5kV, 25mA); C : PLG75:25, l-20um (3kx at 5kV, 25mA); and D: PLG65:35, l-20um (2kx at 7kV, 25mA)]  Figure 17  59  S E M micrographs of 20% paclitaxel-loaded PLG50:50 microspheres [A: I.V.= 0.58, l-20um (lkx at 5kV, 25mA); B: I.V.= 0.74, l-20um (lOkx at 5kV, 25mA); C : I.V.= 0.78, l-20um (2kx at 5kV, 25mA); and D:  I.V.=  1.06, l-20um (3kx at 5kV, 25mA)] Figure 18  60  S E M micrographs of 10% paclitaxel-loaded PLG50:50 microspheres [A: I.V.= 0.58, l-20um (lkx at 5kV, 25mA); B: I.V.= 0.74, l-20um (5kx at 5kV, 25mA); C : I.V.= 0.78, l-20um (5kx at 5kV, 25mA); and D:  I.V.=  1.06, 1-20u.nl (4kx at 5kV, 25mA)] Figure 19  61  S E M micrographs of control PLG50:50 microspheres [A: I.V.= 0.58, 120pm (400x at 5kV, 25mA); B : I.V = 0.74, l-20um (2kx at 8kV, 25mA); C : I.V.= 0.78, l-20pjn (2kx at 5kV, 25mA); and D: I.V.= (5kx at 5kV, 25mA)]  1.06, l-20um 62  XIV  Figure 20  S E M micrographs of paclitaxel-loaded PLG50:50 (I.V.= 1.06, l-20um) microspheres [A: 1% paclitaxel (4kx at 5kV, 25mA); B: 5% paclitaxel, (lkx at 5kV, 25mA); C : 10% paclitaxel, (4kx at 5kV, 25mA); and D: 20% paclitaxel, (3kx at 5kV, 25mA)]  Figure 21  63  X-ray powder diffraction pattern of paclitaxel. Paclitaxel sample as received was scanned at 2 ° / m i n at 20mA and 40kV at room temperature (25°C)  Figure 22  64  X-ray diffraction patterns of A : control and B : 20% paclitaxel-loaded PLG100:0 microspheres. Microsphere samples were scanned at 2 ° / m i n at 20mA and 40kV at room temperature ( 2 5 ° C )  Figure 23  65  D S C thermograms of A : paclitaxel and B: a physical mixture o f paclitaxel/PLG50:50 (I.V.= 1.06, l-20pm) microspheres (paclitaxel: microsphere ratio is 1:4). Samples were scanned at a heating rate o f 10 ° C / m i n with a N 2 purging rate of 40 m L / m i n  Figure 24  70  D S C thermograms of A : 20% paclitaxel-loaded PLG85:15 (I.V.= 0.56, 120um) and B: control PLG65:35 (I.V.= 0.55, l-20um) microspheres. Samples were scanned at a heating rate of 10 ° C / m i n with a N 2 purging rate of 40 m L / m i n  Figure 25  71  Representative F T I R spectra of A : paclitaxel and B: control PLG50:50 (I.V.= 0.74, l-20pm) microspheres. Samples were run at a resolution of 2 cm"  Figure 26  1  73  Representative F T I R spectra of A : a physical mixture o f control PLG50:50 (I.V.= 0.74, l-20um) microspheres/paclitaxel and B: 20% paclitaxel-loaded  XV  PLG50:50 (I.V.= 0.74, l-20u.m) microspheres. Samples were run at a resolution o f 2 cm" Figure 27  1  74  Cumulative amounts o f paclitaxel released from PLG50:50 (I.V.= 1.06, 120u.m) microspheres in PBS-albumin at 3 7 ° C ( N = 4) 82  Figure 28  Cumulative amounts of paclitaxel released from PLG100:0 (I.V.= 0.60, 20100pm) microspheres in PBS-albumin at 3 7 ° C ( N = 4) 83  Figure 29  Cumulative amounts o f paclitaxel released from 10% paclitaxel-loaded P L G microspheres (l-20pm) with different L A : G A compositions in P B S albumin at 3 7 ° C ( N = 4)  Figure 30  84  Cumulative amounts o f paclitaxel released from 20% paclitaxel-loaded P L G microspheres (l-20pm) with different L A : G A compositions in P B S albumin at 3 7 ° C ( N = 4)  Figure 31  85  Cumulative amounts o f paclitaxel released from 10% paclitaxel-loaded P L G microspheres (l-20pm) with a 50:50 L A : G A composition in P B S albumin at 3 7 ° C (N = 4)  Figure 32  86  Cumulative amounts of paclitaxel released from 20% paclitaxel-loaded P L G microspheres (l-20pm) with a 50:50 L A : G A composition in P B S albumin at 3 7 ° C ( N = 4)  Figure 33  87  Cumulative amounts o f paclitaxel released from 10% paclitaxel-loaded PLG50:50 (I.V.= 0.78, l-20um & 20-100um) microspheres in P B S albumin at 3 7 ° C ( N = 4)  88  XVI  Figure 34  G P C elution profiles of A : P L G polymers with the same L A : G A composition (50:50) but different inherent viscosities and B: P L G polymers with different L A : G A compositions but similar inherent viscosities (I.V. ~ 0.55) at 4 0 ° C (Solvent: chloroform. Solvent flow rate: 1.0 mL/min) 91  Figure 35  Degradation profiles of control PLG50:50 microspheres with differing inherent viscosities (l-20u,m). Molecular weights were measured by G P C at 4 0 ° C using polystyrenes as standards (Solvent: chloroform. Solvent flow rate: 1.0 mL/min)  Figure 36  95  Degradation profiles of control P L G microspheres (l-20urn) with different L A : G A compositions and similar inherent viscosities. Molecular weights were measured by G P C at 4 0 ° C using polystyrenes as standards (Solvent: chloroform. Solvent flow rate: 1.0 mL/min) 96  Figure 37  . Degradation profiles of control and 10% paclitaxel-loaded P L G microspheres (l-20u,m) with different L A : G A compositions. Molecular weights were measured by G P C at 4 0 ° C using polystyrenes as standards (Solvent: chloroform. Solvent flow rate: 1.0 mL/min) 97  Figure 38  Degradation profiles of control and 10% paclitaxel-loaded P L G microspheres (l-20u,m) with different L A : G A compositions. Molecular weights were measured by G P C at 4 0 ° C using polystyrenes as standards  XVII  (Solvent: chloroform. Solvent flow rate: 1.0 mL/min) 98 Figure 39  Degradation profiles of control PLG50:50 (I.V - 0.78, l - 2 0 p m & 20100u.m) microspheres. Molecular weights were measured by G P C at 4 0 ° C using polystyrenes as standards (Solvent: chloroform. Solvent flow rate: 1.0 mL/min)  Figure 40  99  S E M micrographs of control PLG100:0 (I.V.= 0.60, 20-lOOum) microspheres degrading in PBS-albumin at 3 7 ° C 101  Figure 41  S E M micrographs of PLG65:35 (I.V.= 0.55, l-20pm) control microspheres degrading in PBS-albumin at 3 7 ° C  Figure 42  S E M micrographs of PLG50:50 (I.V.= 0.58, l-20um) control microspheres degrading in PBS-albumin at 3 7 ° C  Figure 43  102  103  S E M micrographs of PLG50:50 (I.V.= 0.74, l-20um) 10% paclitaxelloaded microspheres degrading in PBS-albumin at 3 7 ° C 104  Figure 44  Glass transition temperatures of control P L G microspheres (I.V. 1.06, 1 20um) with a L A : G A ratio o f 50:50 after 28 days o f degradation in P B S albumin at 3 7 ° C (A: day 1, B: day 3, C : day 7, D: 14 and E : day 28) 105  Figure 45  The degradation profiles in PBS-albumin at 3 7 ° C of non-irradiated control and 20% paclitaxel-loaded PLG85:15 (I.V.= 0.56, l-20pm) microspheres as well as 20% paclitaxel-loaded PLG85:15 (I.V = 0.56, l-20um)  XV111  microspheres after irradiation at a dose of 2.5 Mrad. The slopes of linear regression of the degradation profiles were -0.16 (g/mole/day) for the control microspheres, -0.18 (g/mole/day) and -0.38 (g/mole/day) for the paclitaxel-loaded PLG85:15 microspheres (before and after irradiation respectively) 109  XIX  LIST OF T A B L E S Table 1  L A : G A ratios of the P L G polymers determined using H ' - N M R 50  Table 2  X-ray diffraction results for control and paclitaxel-loaded P L G microspheres. Microsphere samples were scanned at 2 ° / m i n at 2 0 m A and 40kV at room temperature ( 2 5 ° C )  Table 3  66  Glass transition temperatures (T ) of control and paclitaxel-loaded P L G g  microspheres (heating rate: 10 ° C / m i n ; N 2 purging rate: 40 mL/min.) 69 Table 4  Wavenumbers of C = 0 double bond in-plane (8C=0) / out-of-plane (yC=0) deformation absorption of the P L G microspheres (+ paclitaxel). Samples were run at a resolution of 2 cm"  Table 5  Percentages of paclitaxel recovery during extraction from control P L G microspheres/paclitaxel mixtures  Table 6  77  Actual loading of paclitaxel in P L G microspheres with different loading, sizes and L A : G A compositions  Table 8  76  Paclitaxel encapsulation efficiency in P L G microspheres with different loading, sizes and L A : G A compositions  Table 7  72  78  Molecular weights of PLG50:50 microspheres with different inherent viscosities degraded over time in PBS-albumin [molecular weights were measured by G P C at 4 0 ° C using polystyrenes as standards (Solvent: chloroform. Solvent flow rate: 1.0 mL/min)]  92  XX  Table 9  Molecular weights of P L G microspheres with different L A : G A compositions and similar inherent viscosities degraded over time in P B S albumin [molecular weights were measured by G P C at 4 0 ° C using polystyrenes as standards (Solvent: chloroform. Solvent flow rate: 1.0 mL/min)]  Table 10  93  MGPC of P L G microspheres with different L A : G A compositions and 10% paclitaxel loading degraded over time in PBS-albumin [molecular weights were measured by G P C at 4 0 ° C using polystyrenes as standards (Solvent: chloroform. Solvent flow rate: 1.0 mL/min)]  Table 11  94  Glass transition temperatures of control P L G microspheres with L A : G A ratios of 50:50, 65:35, 75:25, 85:15 and 100:0 and different molecular weights after 42-days of degradation in PBS-albumin at 3 7 ° C 106  Table 12  Summaries of the MGPC of non-irradiated control PLG85:15 microspheres and non-irradiated and gamma irradiated (2.5 Mrad) 20% paclitaxel-loaded PLG85:15 microspheres after incubation in PBS-albumin at 3 7 ° C 108  LIST OF ABREVIATIONS  ACN  - Acetonitrile  CIA  - Collagen-induced arthritis  DCM  - Dichloromethane or methylene chloride  DSC  - Differential scanning calorimetry  FTIR  - Fourier transform infra-red spectroscopy  GA  - Glycolide  HFIP  - Hexafluoroisopropanol  I.V.  - Inherent viscosity  IA  - Intra-articular  LA  - Lactide  LS  - Laser scattering Coulter 130  M.W.  - Molecular weight  M  - Number average molecular weight  n  M pc  - Molecular weight determined by G P C  M  - Weight average molecular weight  G  w  NMR  - Nuclear magnentic resonance  0/W  - O i l in water  PBS  - Phosphate buffered saline  PD  - Polydispersity  PLG  - Poly(c?,/-lactide-co-glycolide)  PLG100:0  - Poly(d, /-lactide) with 100% (mole%) lactide  PLG50:50  - Poly(J,/-lactide-co-glycolide) with 50% (mole%) lactide  xxii PLG65:35  - Poly(J,/-lactide-co-glycolide) with 65% (mole%) lactide  PLG75:25  - Poly(d,/-lactide-co-glycolide) with 75% (mole%) lactide  PLG85:15  - Poly(J,/-lactide-co-glycolide) with 85% (mole%) lactide  RA  - Rheumatoid arthritis  SEM  - Scanning electron microscope  Taxol®  - Paclitaxel  THF  - Tetrahydrofuran  X-ray  - X-ray diffractometry  PVA  - Polyvinyl alcohol  XX111  ACKNOWLEDGMENT I would like to thank my supervisor, Dr. Helen Burt for her guidance, patience and encouragement throughout my study at the University of British Columbia. I would also like to thank my committee members: Dr. Gail Bellward, Dr. Colin Fyfe, Dr. Robert Miller, and Dr. Kishor Wasan for their time and guidance. I am very grateful to Dr. Ron Reid for his presiding in my first committee meeting. M y appreciation is also given to people in the Goo Point Lab for their help and encouragement. M y thanks also go to M s . Mary Mager (Department  of Metallurgy)  for her  assistance in using the scanning electron microscope and to people in the Department of Chemistry for allowing me to use their FTIR. A support.  special thank you is given to my wife, Bonnie, for her understanding and  1.  INTRODUCTION Microspheres are fine spherical particles (diameter less than l,000u,m). Upon  encapsulation with a drug, microspheres can be divided into two types: (1) homogenous or monolithic microspheres in which drug is dissolved or dispersed throughout the polymer matrix; (2) reservoir type microspheres in which the polymer matrix surrounds a drug core. Microsphere  drug  delivery  systems using various  kinds  of  biodegradable  polymers have been studied extensively for over 30 years. Long-term controlled releases of growth hormones (Cleland et al, peptides (Mehta et al, agents (Burt et al,  1997), narcotic antagonists (Pitt et al,  1994), local anesthetics (Wakiyama et al,  1981),  1982) and anticancer  1995) have been investigated to eliminate the inconvenience of  repeated injection and for targeted delivery to the site of action. Paclitaxel has been shown to result in the significant regression of existing rheumatoid arthritis (RA) and to prevent the induction of collagen-induced arthritis in animal models (Brahn et al,  1994; Oliver et al,  1994). Intra-articular therapy using anti-  inflammatory steroids is used for patients in whom R A manifests itself in only a limited number of joints. However, the treatment of R A by direct injection of drugs into the joint cavity is limited by the rapid clearance of the drugs from the joint into the blood (Emkey etal, 1996). In this work, paclitaxel was encapsulated into a biodegradable polymer, poly(J,/lactide-co-glycolide) (PLG). Paclitaxel-loaded P L G microspheres in different size ranges, potentially suitable for intra-articular injection, were prepared and characterized. P L G polymers having different compositions of lactide ( L A ) and glycolide ( G A ) , as well as  1  having different molecular weights with the same L A and G A composition, were chosen to study the influence of these factors on the paclitaxel release profiles. The effect of paclitaxel loading in the polymer matrix and the size ranges of the microspheres on in vitro paclitaxel release rates were also assessed. The degradation characteristics of control and paclitaxel-loaded microspheres were determined.  Scanning electron microscope ( S E M )  and differential  scanning  calorimetry (DSC) were used to determine the changes in surface morphologies and glass transition temperatures of the P L G microspheres with or without loading with paclitaxel and before and after incubation in PBS-albumin. Fourier transform infra-red spectroscopy (FTIR) was used to investigate a possible paclitaxel-polymer interaction.  2  2.  BACKGROUND  2.1.  Paclitaxel  2.1.1. Source Paclitaxel is a natural diterpenoid extracted from the bark o f Taxus brevifolia (the Pacific yew). Its efficacy against human cancers was first discovered by Monroe Wall and his collaborators (Wani et al, 1971). Paclitaxel used to be mainly produced by semisynthetic methods from renewable sources such as bark, wood, and dried needles of the yew tree using several different approaches (Magri et al., 1986 & 1988; Denis et al., 1990 and Palomo et al, 1990). A final partial synthetic method was reported by Holton (1990) and was later modified by Ojima et al. (1991) to significantly increase the yield o f paclitaxel. This offered the best overall strategy for the preparation o f paclitaxel from baccatin III and was adopted by Bristol-Myers Squibb (Kingston, 1991). Total synthesis of paclitaxel was also achieved by Nocolaou et al. (1994) and Holton et al. (1994), although this will not be a practical supply source due to the cost.  2.1.2. Chemistry ofpaclitaxel The structure o f paclitaxel, as shown in Figure 1, is "composed of a baccatin III, a typical taxane structure and an ester linked side chain attached at the C-13 position o f baccatin III (Lythgoe et al, 1968). Structurally, paclitaxel is differentiated from most other taxane diterpenoids by its ester side chain at C-13 and its oxetane ring (Kingston, 1991). The 2 ' - O H and 3'-phenyl at side chain (2'R-3'S) are the absolute stereochemical requirements for good binding. The presence of the 4,5-oxetane is necessary for activity but it is not clear whether it is involved in binding or serves as a "conformational lock" on the preferred binding conformation (Suffhess, 1994B).  3  Figure 1  Chemical structure o f paclitaxel (numbers assigned to the  carbons in the structure according to I U P A C nomenclature) (The Merck Index, 1996).  4  The chemical structure and structure-activity relationships of paclitaxel have been reviewed and summarized by Kingston (1991). The most important feature o f paclitaxel's structure with respect to activity is the C-13 side chain. Deletion of the side chain gives baccatin III, which shows a 1700-fold decrease in anti-mitotic activity (Lataste et al, 1984; Miller et al., 1981). A n y modification on baccatin III has been shown to lead to less effectiveness in tumor inhibition (Wani et al, 1971; Miller et al., 1981; Kingston et al, 1982 and Lataste et al, 1984).  2.1.3. Crystal structure X-ray crystallographic data have been reported for paclitaxel precipitated from a dioxane, water and xylene cosolvent (Mastropaolo et al, 1995). Gao et al (1995) also reported the crystal structures o f a paclitaxel solvate and paclitaxel analog (2-debenzoyl, 2-acetoxy paclitaxel).  2.1.4. Solubility and stability Paclitaxel's solubility in various aqueous vehicles and organic solvents such as polyethylene glycol 400, ethanol, isopropanol, soybean oil, acetonitrile and methylene chloride were summarized by Adams et al. (1993). Data from different investigators showed paclitaxel solubility in water at 3 7 ° C was - 3 0 u,g/mL (Swindell et al, 1991; Ringel et al, 1991), -0.7 p g / m L (Mathew et al, 1992), and 6 p.g/mL (Tarr et al, 1987). Recent evidence suggested that the differences in solubility might be due to the presence of different hydrate and anhydrous forms o f paclitaxel (Perrone et al, 1996; Liggins et al, 1997). Dordunoo and Burt (1996) reported that paclitaxel degraded in an aqueous phosphate buffered saline with albumin (PBS-albumin, p H = 7.3). Degradation followed  5  pseudo-first order kinetics and the major degradation products were baccatin III, 10deacetylbaccatin III and their 7-epi isomers (baccatin V and 7-epi-10-deacetylbaccatin III). They also found that degradation of paclitaxel in aqueous buffers was pH dependent and paclitaxel was relatively stable in the pH range of 3 to 5. The degradation products of paclitaxel in methanol at a pH of 9 were found to be similar to those from PBS-albumin (Lataste et al., 1984; Ringel et al., 1987). 2.2.  Paclitaxel in rheumatoid arthritis  2.2.1. Rheumatoid arthritis (RA) R A is a chronic and relapsing inflammatory disease of unknown cause. R A affects the wrist and hand joints, elbows, shoulders, neck, jaw, hips, knees, ankles, and feet in a symmetrical pattern. It tends to persist over prolonged periods of time, and the inflamed joints eventually can become damaged (Figure 2) (Harris, 1997). Harris (1990 & 1997), Hollingsworth et al. (1967) and Lafyatis et al. (1989) have described R A as being characterized by a marked thickening of synovial membrane and the formation of villus projections. The villus projections then extend into the joint space and there is extensive synoviocyte proliferation (or multilayering of the synoviocyte lining) and infiltration of the  synovial membrane with white blood cells (macrophages,  lymphocytes). The tissue formed as a result of this process is called pannus and eventually it grows to fill the joint space. The pannus develops an extensive network of new blood vessels through the process of angiogenesis. Releases of digestive enzymes (such as collagenase and stromelysin) and other mediators of the inflammatory process from the cells of the pannus tissue lead to the progressive destruction of the cartilage  6  tissue. The pannus invades the articular cartilage leading to erosions and fragmentation of the cartilage tissue.  2.2.2. Mechanism of action ofpaclitaxel Paclitaxel induces cytotoxicity by a unique mechanism of action. Paclitaxel affects the fiber-like cell structures called microtubules, which play an important role in cell division and other cell functions. A large number of microtubules are formed at the start of cell division. A s cell division comes to an end, these microtubules are normally broken  down.  However,  paclitaxel  stabilizes  depolymerization back to tubulin (Schiff et al, Suffness,  1994A).  Horwitz  pharmacology of paclitaxel,  and  1979; Howitz et al,  co-workers  demonstrating  microtubules  largely  and  their  1982 & 1992 and  elucidated  that paclitaxel  inhibits  the  molecular  bound preferentially  to  microtubules rather than free tubulin subunits. The binding of paclitaxel to polymerized tubulin is reversible (Schiff et al,  1979 & 1980; Parness et al,  1981 and Manfredi et  al,  1984). Microtubules formed in the presence of paclitaxel are dysfunctional, causing the interruption of normal cell functions, including mitosis (Schiff et al,  1980; Caplow et  al,  1994). In clinical studies, paclitaxel has been demonstrated to be efficacious against many cancers. A 24-hour continuous infusion of paclitaxel (180 to 250 mg/m) gave 20% complete response and 95%> partial response in ovarian cancer patients (Einzig et 1992; Milas et al,  1995). Single agent therapy using paclitaxel gave a response rate of  62%. in breast cancer patients (Arbuck et al, et al,  al,  1994; O'Shaughnessy et al,  1994 and Zoli  1995). In vitro studies have indicated that paclitaxel possesses significant potency  against brain tumors' (Cahan et al,  1994) and bladder cancer (Rangel et al,  1994).  7  Cartilage  Swollen Synovium  Muscle  Normal joint  Figure 2  Inflamed joint  Illustration of normal and inflammatory joints (adapted  from Harris 1997).  2.2.3. Preclinical effectiveness ofpaclitaxel in rheumatoid arthritis Hui et al. (1997) has demonstrated that paclitaxel is capable o f inhibiting synovial cell proliferation  at low concentrations (10" M) and is cytotoxic to  synoviocytes  higher  at  8  concentrations  (10" M). 7  In  contrast,  proliferating  paclitaxel  at  high  concentrations (10" M) has no effect on chondrocyte cell function or viability. Paclitaxel 5  spares non-proliferating synoviocytes and chondrocytes, but is selectively toxic to proliferating synoviocytes which lead to the development o f a locally invasive and destructive pannus tissue. Brahn et al. (1994) used Louvain rats immunized with type II collagen (day 0) to induce arthritis. Paclitaxel was administered beginning on day 2 (prevention protocol) or at the onset o f arthritis on day 9 (in either a high-dose or low-dose suppression protocol). Paclitaxel resulted in significant regression o f existing collagen-induced arthritis and completely prevented the induction of collagen-induced arthritis (CIA). Paclitaxel is particularly cytotoxic for cells undergoing mitosis (Lopes et al,  1993). B y interfering  with normal microtubule function, paclitaxel inhibits cell mitosis (Milas et al, migration  and suppresses intracellular  transport  (Rowinsky  et  al.,  1990).  1995), Hence,  paclitaxel can suppress arthritis because rapidly proliferating, inflammatory pannus cells in R A are more susceptible to paclitaxel's phase-specific cytotoxic effects (Oliver et al., 1994). Oliver and Brahn (1994) suggested that paclitaxel also acted as a phase-specific anti-angiogenesis agent in rheumatoid arthritis.  9  2.3.  Intra-articular therapy  2.3.1. Side effects of paclitaxel systemic therapy Paclitaxel is poorly soluble in water and other common vehicles used for parenteral  administration. Cremophor-EL (polyethoxylated castor  oil) with 50%  dehydrated alcohol (Grem et ai, 1987) has been selected as a vehicle for clinical administration of paclitaxel to cancer patients. A 24-hour infusion schedule with a wide range of paclitaxel dosages (135 to 250 mg/m ) was proposed after the completion of 2  phase I studies (Onetto et al., 1993; Jamis-Dow et al., 1993). However, during phase I and phase II  trials,  serious  side  effects,  such  as  hypersensitivity  reactions,  thrombocytopenia, anemia, arthralgias and myalgias, cardiac effects, alopecia, and gastrointestinal effects (Finley et al, 1994), were reported. The hypersensitivity reactions and cardiotoxicity seemed to be independent of dose (Rowinsky et al., 1992) and were due to the Cremophor E L vehicle used in the formulation (Adams et al., 1993; Rowinsky etal, 1993). 2.3.2. Intra-articular therapy Intra-articular therapy is used for patients in whom the rheumatoid arthritis manifests itself in only a limited number of joints as the systemic therapy will expose the patients to the risk of adverse reactions (Cawley et al., 1969; Itoh, 1992). The treatment of arthritis by direct injection of drugs into the joint cavity is limited by the rapid clearance of the drugs from the joint into the blood (Bird, 1979; Emkey et al., 1996). Using a radio isotope-labeled albumin, Owen et al. (1994) demonstrated that once injected into the synovial cavity, the drugs first diffused into the cartilage, and then were taken up by the cells in the synovium. Finally, the drugs passed  10  through the synovium into the synovial capillaries and then into the systemic circulation (Weiss et al, 1983), or into the lymphatic system. Hunneyball (1986) concluded that the currently available drugs, such as methotrexate (Hoffrneister, 1983) and hydrocortisone, had very short-term (3-4 days) effects in the joints post injection. Microcrystalline steroids such as triamcinolone hexacetonide (TH) possess a relatively prolonged antiinflammatory effect in the joints (See, 1998) because drug release is controlled by the rate o f crystal dissolution. The efficacy and duration o f action o f the microcrystalline steroid preparation may last up to several weeks (Bird, 1979).  2.3.3. Problems associated with intra-articular therapy Hunneyball (1986) and M o w et al. (1991) summarized the principal problems associated with local injection therapy as follows: •  The drug is released into the systemic circulation with consequent production of adverse reactions.  •  Deleterious effects of the drug on cells in the cartilage, ligaments, and tendons due to significant amounts of drugs present in the joints.  •  Local flare reaction due to the physical nature of drug and the risk of infection in the joint with frequent injections.  Because o f these factors, the recommendation is not to inject the joints more frequently than once every three months (Hunneyball, 1986).  2.3.4. Controlled release intra-articular drug delivery systems De Silva et al. (1979) and Dingle et al. (1978) reported the use o f a liposomal  Cortisol palmitate delivery system for rheumatoid arthritis. Intra-articular administration of this liposomal preparation, at a dose equivalent to 2 mg of hydrocortisone, produced a  11  therapeutic effect that was maximal at 48 hours but returned to the pre-injection state after 14 days. Liposomes have also been used to deliver other steroids (Shaw et al,  1979)  or methotrexate  anti-  (Foong et  al,  1993)  for  intra-articular  injection,  but  the  inflammatory effect can only be sustained for 3-4 days.  2.3.5. Controlled release microspheres for arthritis Polymeric microspheres have been investigated for their tolerability in joints and their suitability for delivering anti-inflammatory drugs to the joints. Cavalier et al. (1986) and  Ratcliffe  et  al.  (1984A  &  1984B)  used  polylactic  acid  (PLA),  poly-  butylcyanoacrylate ( P B C A ) , gelatin (PG), and albumin (PA) microspheres as carriers to deliver anti-inflammatory drugs to rabbit knee joints. However, in vivo biocompatibility tests with synovial tissues indicated that P L A , P B C A and P G microspheres with sizes of l - l O p m caused joint inflammation. P A microspheres with sizes o f l-10u,m were well tolerated by the tissues. However, P A microspheres disintegrated rapidly, which resulted in short-lived clinical effects (Ratcliffe et al,  1987).  The use of biodegradable poly(d,/-lactide-co-glycolide) ( P L G ) microspheres with a composition of 85:15  d,/-lactide:glycolide as a 90-day delivery device has been  reported in a U S patent by Tice et al. (1985). A n undisclosed anti-inflammatory agent encapsulated up to 75% (w/w)  in the biodegradable PLG85:15 microspheres, was  administered by intra-muscular injection into the area surrounding the inflamed joints. A continuous release of the anti-inflammatory agent into the joints was achieved.  12  2.4.  Polymer chemistry  2.4.1. Structure Flory (1953) defined polymers as long-chain molecules of very high molecular weights. Depending on their structure, polymers can be divided into two categories, homopolymers and copolymers. A homopolymer is a polymer consisting of many repeating units or so-called structural units, and these repeating units may contain one or more chemically different structures in a regular sequence. A copolymer, however, is a polymer consisting of two or more chemically different units in an irregular sequence (Rabek, 1980).  2.4.1.1.  Configuration and conformation  Polymer configuration refers to the organization of the repeat units along the chain. Configurational isomerism involves the different arrangements of atoms and substituents in a chain that can be inter-converted only by the breakage and reformation of primary chemical bonds (Flory, 1953). Configurational isomers (Jenkins, include  structural  isomers  (such  as  branched/cross-linked  polymers,  1972)  sequence  distributions in copolymers, and head-tail isomers) and steric isomers (cis-trans and tetrahedral such as atactic, isotactic and syndiotactic isomers). Polymer conformation (Flory, 1953) refers to the different arrangements of atoms and substituents of the polymer chain brought about by internal rotation about single bonds. Examples of different polymer conformations (Jenkins, 1972) include the fully extended zigzag, helical, folded chain, and random coils. Polymers can have many possible conformations. For example, a polyethylene without any side chains may possess at least 3 " relatively stable conformations (N is the degree of polymerization). 2N  3  13  2.4.1.2. Morphology and models ofpolymer structures Polymer morphology describes the arrangement, form and structure of polymer chains in crystalline and amorphous regions (Rabek, 1980). Compared to small molecules, polymer morphology is quite complex due to the long chain structures and chain entanglement (Flory, 1953). Polymers are generally considered to be either amorphous or semi-crystalline. The concept of the "Fringed Micelle Model" was first proposed by Herrmann (1930) to describe the nature of semi-crystalline polymers (Figure 3A). In the "Fringed Micelle Model", crystallites are interspersed in an amorphous matrix. Some polymer chains contribute to both the amorphous and crystalline regions in the polymer matrix. Amorphous and crystalline regions co-exist and are not separated by distinct boundaries. However, the "Fringed Micelle Model" fails to explain the observation that polymer single crystals are generally 100 A in thickness and are much shorter than the length of the polymer chains. Keller (1957) later used a "Chain Folded Model" (Figure 3B) to show how polymer chains could fold and form a single crystal without changing the bond angle. In this model, lamellae are formed by adjacent re-entry of a single polymer chain. In contrast, Flory (1962) proposed a random reentry model or "Switchboard Model" (Figure 3C). He believed a single polymer chain could enter the same lamellae or different lamellae randomly at different location. The discovery of inter-lamellar links between poly(ethylene) lamellae by Keith et al. (1966) supported the switchboard model.  14  Amorphous regions  Crystalline regions  A  Fringed micelle model  Folded chain model  Switchboard model Figure 3  Fringed-micelle,  (adapted from Rabek, 1980).  folded chain and switchboard  models  Amorphous polymers do not exhibit either a crystalline X-ray diffraction pattern or a first order melting transition (Rabek, 1980). Irregularity in the polymer chain, chain entanglement, branching, and cross-linking can prevent polymer crystallization when polymers are cooled from the liquid state. However, some semicrystalline polymers can behave as amorphous polymers i f they either have a very slow crystallization rate (such as polycarbonate and polyethylene terephthalate) or only become crystalline at very low temperatures (such as natural rubber at - 2 5 ° C ) . Flory (1953) believed that amorphous polymers did not have any long-term and short-term order in the matrix and could be described by his "Single-phase Random-coil M o d e l " (Figure 4A).  This model gives an  excellent explanation of the structures of elastomers, but fails to interpret the nodule structures in many amorphous polymers and the much higher X-ray diffraction intensity in molten polymers. Y e h (1972) later solved this problem by introducing a "Folded-chain Fringed Micellar Grain M o d e l " (Figure 4B). According to his model, amorphous polymers can be divided into "grain" (G) and "intergrain" (IG) regions and the "grain" is composed of "grain boundary" (GB) and "ordered domain" (OD).  16  A  B  Figure 4  Single-phase random-coil model  Folded-chain fringed micellar grain model  Single-phase random-coil and folded-chain fringed micellar grain models  (G: grain; IG: intergrain; G B : grain boundary and O D : ordered domain) (adapted from Sperling, 1986).  17  2.4.2. Molecular weight Physicochemical properties of polymers are determined  not  only by  their  chemical structures but also by their molecular weights. Polymers consist of molecules which have a distribution of molecular weights instead of possessing a single molecular weight  like  small molecules (Flory,  1953).  Several  important  molecular  weight  parameters are used to describe polymer systems. The most common ones are the number-average molecular weight (M ), which is defined as the total weight of all solute n  species divided by the total number of moles present (equation 1) and the weight-average molecular weight ( M ) , which is defined in equation (2): w  y  y w,  NM,  M . = ^ =  £  y  = = ^  X  N,  y  NM,  2  M „ = 4^  £  Equation 1 W  i / M l  WM,  =  Equation 2  N,M,  W  1  Where: M;, Nj and Wj are the molecular weight, number of moles, and weight of the ith solute species, respectively. Both M  n  and M  w  are unique values for a given polymer. M  is most affected by the lower molecular weight fraction, whereas M the higher fraction. Therefore M  w  is always greater than M  n  w  n  is most affected by  (Figure 5).  A single number or index has also been used to describe the molecular weight distribution of a polymer. The extent of variability in molecular weight is determined by the polydispersity (PD), which can be calculated from the following equation:  M  PD = — -  M.  Equation 3  18  Figure 5  Molecular weight distribution of a polymer (Rabek, 1980).  19  2.4.3. Biodegradable polymers A variety o f synthetic polymers have been reported to degrade and erode in mammals. The degradation products o f these polymers can be eliminated from the body by either metabolism or renal filtration.  Polymers that undergo degradation in a  biological environment, through either simple chemical reactions or enzyme-catalyzed reactions, are designated as biodegradable polymers (Vert et al., 1991). These polymers can function as a matrix to control the drug release by polymer hydration and degradation, which is then followed by polymer bulk erosion and elimination o f the degradation products from the body (Park et al., 1996). A great number o f biodegradable polymers have been developed and evaluated for drug delivery, but the most widely investigated biodegradable polymeric systems are the lactide:glycolide polymers.  2.4.4. Glass transition 2.4.4.1.  Glass transition temperature (Tg)  The glass transition is the change o f an amorphous polymer or region from a hard, brittle, glassy state to a soft, flexible, rubbery state or vice versa (Figure 6). The temperature o f this transition is called the glass transition temperature (T ) (Turi, 1997). g  Below the glass transition, chain segments are frozen in fixed positions. A diffusionrearrangement  o f the segmental position is also less probable.  With increasing  temperature, the amplitude of segmental vibrations increases. In the transition-state, chain segments have sufficient energy to overcome the secondary bonding forces. Chain segments or chain loops may perform rotational and translational motions and the polymer becomes flexible (Flory, 1953).  20  The glass transition temperature is well marked in amorphous polymers. In semi-crystalline polymers, it is less conspicuous because it only occurs in the noncrystalline regions. 2.4.4.2. Factors influencing glass transition temperature A  decrease in  chain flexibility  due to rigid  and bulky  side groups and  intermolecular bonds such as cross-linking or hydrogen bonds will lead to an increase in T . Factors leading to a decrease in T include end group effect (branching) and presence g  g  of low molecular weight compounds, such as water, solvent and plasticizers (Rabek, 1980). In semicrystalline polymers, T  g  increases proportionally with the degree of  crystallinity, because the crystallites tend to reinforce or stiffen the structure. The T of most copolymers falls between the glass transition temperatures of their g  individual polymers. A n amorphous random copolymer exhibits a single T which can be g  predicted by:  T  Equation 4  =  KW +W  8  l  2  Where T i and T 2 are the glass transition temperatures of two homopolymers. W i and g  g  W 2 are the weight fractions of two homopolymers and K is a constant given by: K = (a  r  - a  g  ) l(a 2  r  -a )  g x  Equation 5  Where cci and ct2 are the thermal expansion coefficients of the homopolymers, the subscript r referring to the rubber state and the subscript g referring to the glassy state. The molecular weight dependence of the glass transition has been defined by the Fox-Flory equation (Fox et al., 1950): g  T  =  s^)  T  ~™«  1  Equation 6  21  Where T (oo) is the limiting high-molecular weight value of T g  g  and K is a constant related  to free volume.  Exothermic behaviour  Crystallization Second-order or glass transition  Curing, oxidation,chemical reactions.orcrosslinking  is  o o  k 0 i <  Shifted base Solid-solid first-order transition  Melting transition I  DTA curve  Figure 6  Degradation or vaporization  DSC curve \  «*——i  D S C thermogram for a polymer illustrating the different,  possible thermal events (Rabek, 1980).  22  2.5.  Poly(tf,/-lactide-co-glycolide) ( P L G )  2.5.1. Structure and synthesis L o w molecular weight P L G (usually less than 10,000 g/mole) is prepared from the direct copolycondensation of (d,l)  lactic acid and glycolic acid, with or without a  catalyst such as antimony, or tin at 180°C and 4 mmHg. P L G with higher molecular weight is manufactured by the ring opening melt-polycondensation o f lactide and glycolide, the cyclic dimers of lactic acid and glycolic acid respectively, at higher temperatures ( 2 2 0 ° C and 0.1 mmHg) in the presence of the same kinds of catalysts (Figure 7). This process gives a higher molecular weight ranging from 20 to 250 kg/mole (Gilding et al, Unlike  1979; Reed et al, glycolide, which  1981). contains only  symmetric  atoms, lactide  has  two  asymmetric carbon atoms and is thus a chiral molecule that exists as two optical isomers or enantiomers, ^-lactide and /-lactide and a racemic d, /-lactide. The lactide isomers give rise to four morphologically distinct polymers: poly(J-lactide) and poly(Mactide), which are two stereoregular polymers; and poly(meso-lactide) and po\y(d,/-lactide),  which are  racemic polymers. Polymers derived solely from c?-lactide, /-lactide and glycolide are semicrystalline. The homopolymers of d,/-lactide  and copolymers o f d,/-lactide  glycolide are all amorphous polymers (Gilding et al,  1979; Fukuzaki et al,  and  1991).  23  Sn 2+ O"  "0  CH  + 3  </, / - lactide  H  0.1 mmHg  CT  200°C  Glycolide  CH CH I I -c—C—O—C—C—O—]— I II I II  H  3  - C — C— O— C— C— O — ]„ II I II ' H O H O  3  m  H O  (GA)  H O  (d, /-LA) A . Synthetic pathway for P L G polymers  O  CH CH 2  3  I I Sn  0 — C — CHCH CH CH2 CH 2  O—  C— II  2  3  CHCH2CH2CH2CH3  I  O  CH CH 2  3  B. Chemical structure of stannous 2-ethyl-hexanoate catalyst  Figure 7  Synthetic  pathway  for  P L G polymers  and chemical  structure of stannous 2-ethyl-hexanoate catalyst (Zhang et al., 1994).  2.5.2. Drug delivery systems (DDS) based on PLG 2.5.2.1. Microspheres P L G microspheres have been successfully used in the formulation o f many pharmaceutical agents as summarized by Ilium et al. (1982); Holland et al. (1986 & 1992) and Lewis (1990). Upon encapsulation o f a drug, microspheres can become a monolithic type microspheres in which the drug is dissolved or dispersed throughout the polymer matrix or a reservoir-type microspheres in which the polymer matrix coats the drug particle.  2.5.2.2.  Biocompatibility  Outright et al. (1971) and Craig et al. (1975) first demonstrated the biocompatibility of P L G using sutures. When implanted into rats, these polymer sutures induced a mild local inflammatory  reaction, infiltration of macrophages, giant cells and fibroblasts. The  sutures were surrounded by a thin layer o f connective tissue until they were completely absorbed. Visscher et al. (1985) also reported the tissue response after intra-muscular injection  o f P L G microspheres containing  lysine-8-vasopressin and described the  morphologic changes in the microspheres based on histological observations. In the case of a device with a one-month life-span, a minimal inflammatory reaction characterized by infiltration o f lymphocytes, plasma cells, histocytes, and acute myositis was observed soon after injection. B y day 63, degradation and erosion o f the microspheres were extensive  and the minimal  chronic cell response had almost completely  resolved  (Visscher et al., 1985). Spenlehauer et. al. (1989) also observed that at about 6 weeks after implantation, the absorption o f microspheres was almost completed. The intensity of the inflammatory response had decreased and only remnants o f microspheres could be  25  seen. K o u et al. (1997) reported that P L G rods and their breakdown products were well tolerated by the brain tissue upon implantation into rat brains. In general, the tissue response to P L G is a very localized inflammation consisting o f macrophages, foreign body giant cells, and capillary infiltration and it is generally accepted that P L G polymers are biocompatible with living tissue (Maulding, 1987).  2.5.2.3. Mechanism of in vitro degradation and erosion According to traditional definitions, only chemical agents can cause chemical degradation o f polymers and biodegradation is caused by enzymes, bacteria, and fungi. Graham and W o o d (1982) defined biodegradation as the process in which polymers degraded after a period of time to soluble species, which were readily removed from the in vivo site and excreted by the body. We use the term "degradation" or "biodegradation" to describe the chain scission process throughout the polymer matrix. During degradation, polymer chains are cleaved to form oligomers and finally to form monomers. Degradation may not involve any mass loss during the process (Gopferich, 1996). Erosion, on the other hand, is the process o f conversion o f an initially water-insoluble material to a water-soluble material with an accompanying mass loss (Heller, 1980 & 1985). Erosion emphasizes the loss of material owing to monomers and oligomers leaving the polymer surface (surface erosion) or matrix (bulk erosion) (Tamada, et al, 1993). P L G polymers are a class o f biodegradable polymers suitable for drug delivery systems since they are biocompatible with living tissue and the non-toxic degradation products are readily removed from the site o f application (Holland et al., 1992). P L G is  26  degraded via a labile ester bond. It also has suitable mechanical properties and can be sterilized with limited impairment of properties. The degradation of P L G is a random process which is characterized by rupture of the P L G main-chain. P L G used in drug delivery systems is generally degraded by hydrolysis. Water molecules diffuse into the amorphous polymer matrix, followed by random hydrolytic cleavage of ester bonds to produce oligomers. Erosion can commence with the degradation process (Heller, 1984). In many cases, the release rate of the encapsulated drug depends on drug diffusion rate and the rate of degradation and erosion of P L G polymers (Lewis, 1990). During degradation, the molecular weight of the polymer steadily decreases, but the matrix can remain in its original shape and retain mass until a critical molecular weight, when bulk erosion begins (Iwata et al., 1993).  2.5.2.4.  Factors affecting PL G degradation  Factors affecting the degradation and erosion of P L G include p H , molecular weight, and temperature (Reed et ai,  1981; Vert et al,  1991). A discontinuity of mass  loss for P L G is showed at the glass transition temperature (T ) in which the mass loss g  above T  g  is greater than that below T , because water uptake and hydrolysis occur more g  readily when the temperature is above T . The molecular weights o f the P L G also g  profoundly affect the degradation kinetics. The degradation rate is decreased for lower molecular weight P L G polymer (Reed et al,  1981).  P L G polymers possess different compositions of lactide and glycolide, which provide for different life-spans in living tissue. The half-life of P L G polymers with various compositions of lactide:glycolide, molecular weights and crystallinity, generally  27  ranged from 2 weeks to 6 months. A n amorphous poly(Mactide-co-glycolide) with 70:30 mole% and weight average molecular weights o f 16,900-41,300 g/mole showed 65% mass loss after 10 weeks o f implantation (Fukuzaki  et al., 1991). In in vivo studies,  Maulding (1987) and Visscher et al. (1985) showed that a 50:50 mole% P L G with a number-average molecular weight o f 9,500 g/mole was almost completely resorbed in tissue by day 63.  2.5.2.5. Mechanism of drug release from PLG matrix In vitro drug release from P L G microspheres generally follows a triphasic profile (Sanders et al., 1986). Initially, a burst phase is observed over the first few days due to the release o f drug from the surface o f the microspheres and the matrix located near the surface. There is a period o f slower release when the degradation medium diffuses into the polymer matrix. Degradation begins and at the same time, the drug diffuses slowly out of the matrix. The third phase, a secondary burst phase o f drug release, occurs when the bulk erosion of polymer matrix begins. In some cases, the third phase does not appear due to significant amount o f drug being released in a short period o f time, leading to a biphasic release profile.  2.5.2.6. Factors affecting drug release from PLG matrix Factors affecting the degradation and erosion of P L G microspheres also affect the release rates o f drug from the polymer matrix. Physical properties o f polymers, such as the molecular weight and the composition (the ratio o f the more hydrophobic lactide fraction to the less hydrophobic glycolide fraction),  have a major influence on the drug  28  release since polymer degradation and erosion are dependent on these factors (Sanders et al, 1986; O'Hagan et al, 1994 and Ruiz et al, 1990). The sizes of the microspheres may be important in the early phase of drug release. Since smaller microspheres have a larger surface area, the release rate may be faster. The molecular weight and structure of drugs encapsulated into the polymer matrix also play a role in release kinetics. The diffusivity of larger molecules through a polymer matrix is lower (Holland et al, 1986) and therefore slower drug release from the matrix is expected. Increasing drug loading by creating a network o f pores within the matrix provides more pathways for drug release. A n extensive pore network will facilitate rapid elution of the drug from the matrix (Takada et al, 1994).  2.5.3. Enzymatic degradation The role o f enzymatic attack in the degradation o f P L G is less defined. Poor in  vivo - in vitro correlation in P L G degradation rates has often been taken as evidence o f enzymatic involvement. For polymers used below their glass transition temperature (T ), g  enzymatic degradation is not thought to be a major factor during matrix breakdown. Enzymes may play a more significant role in P L G degradation above the polymer T  g  (Holland et al, 1992).  2.5.4. Effects of sterilization on microspheres Presently, the most expedient method for sterilizing moisture and heat sensitive substances such as P L G is Co-60 gamma irradiation. Prior studies demonstrated that gamma irradiation of P L G induced dose-dependent chain scission and molecular weight loss (Hausberger et al, 1995). However, it has been reported that the degradation due to irradiation caused no significant change in the initial drug release rate in the dose range  29  up to 2.5 Mrad. The decrease in glass transition temperature caused by irradiation is not large enough to affect the initial release rate (Yoshioka, et al., 1995).  2.5.5. Paclitaxel-loaded PLG microspheres In this work, paclitaxel-loaded microsphere formulations have been developed for intra-articular delivery o f paclitaxel based on P L G . Formulations of paclitaxel-loaded microspheres  suitable  for  intra-articular  administration  have  not  previously  been  developed. Furthermore, the size of microspheres suitable for intra-articular injection has not been determined. PLG  microspheres containing paclitaxel  were  prepared  using the  solvent  evaporation technique. Microspheres were extensively characterized and in vitro release rates of paclitaxel determined. The effects of different molecular weights of P L G polymers and differences in lactide:glycolide compositions in the polymers on the physicochemical characteristics and release properties of the microspheres were assessed.  30  2.6.  Objectives The overall objective of the research was to prepare and characterize paclitaxel  loaded microspheres using the biodegradable lactide:glycolide polymers, which might potentially be suitable for the intra-articular delivery o f paclitaxel in rheumatoid arthritis.  Specific Aims The specific aims of the proposed research were to: •  Prepare  and characterize paclitaxel-loaded microspheres using different  molecular weights and lactide:glycolide ratios of P L G polymers. •  Determine the degradation behavior of the polymers.  •  Determine the in vitro release rates of paclitaxel from P L G microspheres.  31  3. 3.1.  EXPERIMENTAL Materials & supplies  3.1.1. Lactide:glycolide polymers P L G with different inherent viscosities (I.V.) and lactide:glycolide compositions were purchased from Birmingham Polymers Inc. (Birmingham, A B ) . Polymers were stored in a desiccator filled with nitrogen at room temperature as required.  3.1.2. Paclitaxel Paclitaxel was purchased from Hauser C o . (Boulder, C O ) and stored in a refrigerator at - 4 ° C as required.  3.1.3. Chemicals & solvents Dichloromethane ( D C M ) , H P L C grade, Fisher Scientific (Nepeon, O N ) . Poly(vinyl)  alcohol ( P V A ) , M.W.= 13,000-23,000 g/mole, 98% hydrolyzed,  Aldrich (Milwaukee, WI). Methanol, chloroform, and acetonitrile, H P L C grade, all from Fisher Scientific (Fairlawn, N e w Jersey). Polysorbate 80, Sigma (St. Louis, M O ) . Potassium bromide (KBr), F T I R grade, from Sigma (St. Louis, M O ) . Bovine serum albumin Fraktion-V (Boehringer Mannheim, German). Sodium  phosphate  mono-hydrogen  (Na HP04) 2  and  sodium  dihydrogen  orthophosphate ( N a H P 0 H 0 ) , all from B D H Inc. (Toronto, O N ) . 2  4  2  Sodium chloride (NaCl), Acros Organics, (Beljium, N e w Jersey). Acetone, analytical grade, Fisher Scientific (Fairlawn, N e w Jersey). Ethanol, analytical grade, Fisher Scientific (Fairlawn, N e w Jersey).  32  D S C indium standard (99.99%), Sigma (St. Louis, M O ) . Chloroform-D (99.8%) from Cambridge Isotope Laboratory ( M A ) . Distilled water.  Narrow polystyrene (PS) standards with molecular weights o f 2,000, 4,000, 9,000, 17,500, 30,000, 50,000, 100,000 and 170,600 g/mole from Pressure Chemical Company.  3.1.4. Buffers, mobile phases and PVA solution Phosphate buffered saline with 0.4 g/L albumin (pH = 7.3) was prepared with 2.60 g o f sodium phosphate mono-hydrogen  (Na2HP04), 0.32 g o f sodium dihydrogen  orthophosphate ( N a H P 0 H 0 ) , 8.22 g of sodium chloride (NaCl), and 0.4 g of albumin. 2  4  2  A l l above chemicals were then dissolved in 1.0 L distilled water. H P L C mobile phase was prepared by mixing 50 m L o f methanol, 370 o f m L distilled water, and 580 m L of acetonitrile. A l l solvents were H P L C grade. P V A solution (5.0% (w/v)) was prepared with 50 g o f P V A dissolved in 1.0 L distilled water at a temperature of no higher than 7 5 ° C .  3.1.5. GPC analysis A l l G P C standards and polymer sample solutions were prepared by dissolving the corresponding polymer in chloroform ( H P L C grade). For G P C standards and polymer samples with molecular weights less than 20,000 g/mole, a 0.20% (w/v) solution was used. For G P C standards and polymer samples with molecular weights greater than or equal to 20,000 g/mole, a 0.10% (w/v) solution was used. Pure chloroform ( H P L C grade) was used as G P C mobile phase.  33  3.1.6. Glassware Release studies used Pyrex® brand test tubes (15 m L ) with Teflon®-lined screwcap lids. Beakers (250 m L ) and graduated cylinders used for manufacturing microspheres were also Pyrex® brand. Scintillation vials (20 m L ) with polypropylene-lined screw-cap lids were used to store microspheres. A l l glassware was purchased from Fisher Scientific (Toronto, ON).  3.2.  Equipment  3.2.1. Apparatus for microsphere manufacture The  apparatus  for microsphere  manufacturing  consisted o f a Dyna-Mix  overhead stirrer with a speed controller (model-143, Fisher Scientific, Rockford, IL). A 4-blade impeller with shaft length = 203 m m and shaft diameter = 6.4 m m (each blade was a 9 0 ° sector which was 15° inclined from horizontal) was immersed in a 250-mL beaker containing the P V A solution which was emulsified by stirring (Figure 8).  3.2.2. High performance liquid chromatography (HPLC) Chromatographic analyses o f paclitaxel were performed using a Shimadzu system (Tokyo, Japan), equipped with a Shimadzu S I L - 9 A autosampler (Tokyo, Japan), a Shimadzu model S P D - 6 A detector (Tokyo, Japan), a Shimadzu Chromatopac C - R 3 A integrator (Tokyo, Japan), and a Beckman 11 OA pump from Beckman Instruments Inc. (Palo alto, C A ) . The analytical column was a C-18 reverse phase column from Polymer Sciences Inc. (Boston, M A ) .  3.2.3. Gel permeation chromatography (GPC) G P C analyses o f polymers were performed using a Shimadzu system (Tokyo, Japan), equipped with a Shimadzu S I L - 9 A autosampler (Tokyo, Japan), a Shimadzu  34  refractive index detector R I D - 6 A (Tokyo, Japan), a Shimadzu Chromatopac C-R601 integrator (Tokyo, Japan), and a Shimadzu L C - 1 0 A D pump  (Tokyo,  Japan). The G P C  column was a PLgel column with a nominal pore size of 10 A, bead size of 5 um and 7.5 4  x 300 m m in dimensions (Polymer Laboratory, Boston, M A ) .  3.2.4. Differential scanning calorimetry (DSC) Thermal analysis was performed using a differential scanning calorimeter, model 91 OS from Dupont Instruments Inc. (New Castle, D L ) . The thermal analysis system was controlled by an I B M compatible computer loaded with a thermal analyst software from T A Instruments.  Aluminum sample pans and lids were from Rheometric  Scientific  (Piscataway, NJ) and were sealed with a crimper (Dupont, model 900878-901).  3.2.5. Scanning electron microscopy (SEM) Microsphere morphology was determined  using a Hitachi  S-2300 scanning  electron microscope (Tokyo, Japan). Samples were coated with gold-palladium using a Hummer sputter coater (Technics, Alexandra, V A ) .  3.2.6. Fourier transform infra-red spectroscopy (FTIR) FTIR  measurements  were  conducted using  a Fourier  transform  infra-red  spectrometer (Bomem Inc., Quebec). K B r sample discs were compressed using a F T I R K B r sample die (Wilmad, NJ).  3.2.7. Particle size analysis The size distributions of microspheres were measured with a Coulter L S I 3 0 laser scattering particle size analyser (Coulter Scientific, Hialeah, F L ) with Coulter L S I 3 0 version 1.53 computer software.  35  3.2.8. Polymer compositions The molar compositions o f lactide:glycolide polymers were measured using nuclear magnetic resonance ( N M R ) , Bruker Instrument A C - 2 0 0 (Bruker, Germany).  3.2.9. X-ray powder diffraction Rigaku Geigerflex X-ray diffractometer (Tokyo, Japan) was used to determine polymer and paclitaxel X-ray powder diffraction patterns.  3.2.10. Centrifuges A centrifuge, model G P R (Beckman Instruments Inc., Palo Alto, C A ) was used during microsphere manufacture.  A high-speed centrifuge,  model G S - 6  (Beckman  Instruments Inc., Palo Alto, C A ) was used for in vitro paclitaxel release studies.  3.2.11. Incubator and ovens A n isotemp incubator (Fisher Scientific, Fairlawn, NJ) and a culture tube rotator with a rotation speed control unit ( V W R , Toronto, O N ) were used for microsphere release studies. Napco vacuum oven model 5831 (Precision Scientific, Chicago, IL) equipped with an Emerson vacuum pump model S A 5 5 N X G T E 4 8 7 0 (Emerson Motor, St. Louis, M O ) was used to dry paclitaxel and K B r for F T I R studies.  3.2.12. Other equipment Olympus optical microscope model B H - 2 (Olympus Optical, Japan). Contax 35 m m camera, model 167MT (Kyocera Corp., Tokyo, Japan). Mettler balances model PJ300 (Mettler Instruments, Zurich, Switzerland). Vortexer ( V W R , Bohemia, N Y ) . Reacti-Therm III heating/stirring module (Pierce Inc. Rockford, IL). Refrigerator & freezer (Caltel Scientific, Richmond, B C ) .  36  A : Dyna-Mix overhead stirrer with a speed controller  pipette  B: 4-blade impeller  Figure 8  Illustration  of  microsphere  fabrication  (A:  overhead stirrer with a speed controller and B: 4-blades  Dyna-Mix overhead  impeller).  37  3.3.  Methods  3.3.1. Preparation of microspheres Microspheres in the size range o f l-20u.m were prepared using the solvent evaporation method (Jeffery et al., 1991; O'Hagan et al, 1994). A total weight o f 0.50 g of polymer and paclitaxel were dissolved in 10 m L o f D C M . The weights o f paclitaxel used ranged from 5 mg to 100 mg, depending on the paclitaxel loading in the polymer matrix. A 5% (w/v) P V A aqueous solution (100 m L ) was added into a 250-mL beaker using a graduated cylinder. The P V A solution was stirred with an overhead impeller at 900 ± 5 0 rpm at room temperature. The paclitaxel/polymer solution was slowly added into the P V A emulsion. After 2.5 - 3 hours, the suspension o f microspheres was centrifuged at lOOOxg for 10 minutes with a bench top centrifuge. The supernatant was removed by suction and the microspheres were washed four times with distilled water and then  centrifuged.  The washed microspheres were transferred  into  a  20-mL  scintillation vial and air-dried overnight. The dried microspheres were stored at room temperature in a desiccator for further drying. Control microspheres  (paclitaxel-free)  were prepared the same way as described above. Microspheres with size ranges between 20-100pm were prepared by altering the P V A concentration and stirring rate. The paclitaxel/polymer solution was slowly added into 100 m L 2.5% (w/v) P V A aqueous solution and stirred with an overhead impeller stirrer at 550 ± 50 rpm at room temperature. After 2 . 5 - 3 hours, the suspension of microspheres was centrifuged at 170 x g for 10 minutes. The supernatant was removed and the microspheres were washed four times with distilled water. The washed  38  microspheres were then  air-dried  overnight  and stored in a desiccator at room  temperature for further drying. Control microspheres (paclitaxel-free) were also prepared the same way as described above. The yields of all microspheres were calculated.  3.3.2. Characterization of microspheres 3.3.2.1. Nuclear magnetic resonance The molar  compositions o f the polymers were characterized  by  H^NMR  spectroscopy. Briefly, 6-10 mg o f each pure polymer was placed in a seven-inch round bottom N M R tube (I.D.= 5 mm) and a 4.5 cm height of high purity CDCI3 solvent was added into the tube to completely dissolve the samples. The spectra were recorded on a Bruker A C - 2 0 0 (200 M H z ) N M R spectrometer using tetramethylsilane  ( T M S ) as an  internal reference.  3.3.2.2. Measurement of molecular weight by gel permeation chromatography Polystyrene (PS) standards or polymer sample solutions for G P C testing were obtained by dissolving 1.0 to 2.5 mg o f each corresponding polymer in 1.0 m L o f chloroform at room temperature.  Each sample solution was injected into the G P C 4  equipped with a PLgel column (nominal pore size o f 10 A ) through a loop injector (20 u.L). The instrument was stabilized at least one hour before measurement and the temperature was controlled at 4 0 ° C throughout the testing. The solvent flow rate was 1.0 mL/min. The calibration curve o f polystyrene standards was obtained by plotting logarithms of the molecular weights of polystyrene standards versus the retention time.  39  3.3.2.3. Determination of sizes and morphology of microspheres A.  Optical microscope For dried microspheres, 5 mg of microspheres was reconstituted with 0.5 m L of  distilled water and 0.5 m L of 1% of polysorbate 80 solution. One drop of reconstituted microsphere suspension was placed on the glass slide and photographed.  B.  Laser scattering The  microsphere  size  distributions  were  determined  by  Coulter  LSI30.  Approximately 5 mg of microspheres were resuspended in 2.0 m L of 1.0% polysorbate 80 solution. When a background scan on distilled water was completed, 1.0 m L of the above microsphere suspension was added into the sample cell (obscuration reading should be between 7 and 11%). The sample was measured by two 60-seconds scans. A volume% against particle size distribution was obtained along with sample mean ± standard deviation.  C.  Scanning electron microscope The microsphere samples were sprayed on to the conductive adhesive film on the  sample holders. The sample holders were then placed into the sample coater connected with a high vacuum. When a vacuum of 10" torr was reached, argon gas was turned on 5  and a high voltage was applied to provide a current of about 25 m A . Samples were coated with gold-palladium for 3 minutes (about 100 A in thickness). A second coating of the microsphere samples was done at an angle of 4 5 ° . After coating, samples were placed into the S E M (Hitachi model S-2300) to observe the surface morphology at 25 m A and less than 10 k V (with most favorable condition at 5-7kV). Pictures were taken using a computer interface (PCI3.0).  40  3.3.2.4. X-ray diffraction Powder X-ray diffraction analysis on control and paclitaxel-loaded microspheres was conducted with a Rigaku Geigerflex powder X-ray diffractometer. Between 0.7-1.0 g of microsphere sample was pressed firmly in the X-ray sample frame. A n empty sample frame was used when doing the background scan. A l l samples were scanned from 5 to 50° (29) at a rate o f 2 ° / m i n . Pure paclitaxel as received from the manufacturer was scanned in the same way. The X-ray source was C u K « i radiation (k = 1.5418A) and was generated at 40 k V and 20 m A .  3.3.2.5. Thermal analysis The glass transition temperatures (T ) of microspheres with or without paclitaxel g  loading were measured with a Dupont model 91 OS differential scanning calorimeter. About 4.50 ± 0.20 mg samples were accurately weighed in a D S C pan and then sealed with a D S C lid with a Dupont sample crimper. A n empty pan and lid was also crimped and used as a reference pan. A l l samples were analyzed at a heating rate o f 1 0 ° C / m i n with N  2  gas flowing at a rate o f 40 m L / m i n through the sample cell. The D S C was calibrated with indium each time before running samples. Indium  was weighed (4.4 mg) and sealed in the D S C pan and run under the same conditions as the samples. The standard melting temperature for indium is 156.7°C.  3.3.2.6. Fourier transform infra-red spectroscopy (FTIR) F T I R was used to investigate a possible interaction between the paclitaxel and P L G polymers. Briefly, K B r was heated to 4 0 ° C and carefully ground under an infrared lamp to avoid condensation of atmospheric moisture. The K B r was dried at 105°C for 1 hour under a nitrogen atmosphere and then stored in a vacuum oven at room temperature.  41  Both K B r powder and paclitaxel were dried in the oven at room temperature and at reduced pressure (20 mmHg) overnight. P L G microspheres (2.00 ± 0.02 mg), with or without paclitaxel, were mixed with 200 ± 1 . 0 mg of ground K B r . Then 150 ± 2.0 mg of mixed samples was immediately placed in a K B r die. The samples were pressed under initial pressure (2 tons) for 15 minutes and then under high pressure (8-10 tons) for another 15 minutes to form a transparent K B r disc (with a diameter of 13 m m and 0.3 m m in thickness). The sample discs were loaded in an F T I R sample disc holder. A background scan was done to calculate the adsorption due to air and then the samples were scanned between 400 - 4000 cm" with a resolution of 2 cm" . The final spectrum 1  1  was obtained by subtracting the background adsorption. T o obtain a spectrum of a physical mixture, 0.20 ± 0.02 mg of paclitaxel and 1.80 ± 0.02 mg of microspheres (paclitaxel-free) were mixed with 200 ± 1 . 0 mg of ground K B r . Sample discs were prepared and tested as described above. A pure paclitaxel F T I R spectrum was obtained using the same conditions as the microsphere samples. Briefly, 1.00 ± 0.02 mg of dried paclitaxel was mixed with 200 ± 1.0 mg of pre-dried K B r powder. Sample discs were prepared and tested as described above.  3.3.3. Total content of paclitaxel in microspheres 3.3.3.1. Determination of total content ofpaclitaxel in microspheres The assay for measuring total content of paclitaxel in microspheres was based on the method of Burt et al. (1995) with slight modification. Approximately 2 mg of paclitaxel-loaded microspheres, accurately weighed, were dissolved in 1 m L of D C M in a test tube and the resulting solution was transferred to a 100-mL volumetric flask. Most of  42  the D C M was allowed to evaporate at room temperature and 100 m L o f acetonitrile was then added to the flask. A portion of the A C N solution (5 m L ) was then added to a clean test tube and the residual D C M was further evaporated by vortexing for 2 minutes. The resulting solution was centrifuged and 20 u.L of each sample was injected into the H P L C . The samples were analyzed for paclitaxel using H P L C  with a mobile phase o f  water:methanol:ACN (37:5:58) at a flow rate o f 1.0 m L / m i n and U V detection at 232 nm (Burt et al., 1995). The paclitaxel concentration was calculated from the standard calibration curve. Paclitaxel encapsulation efficiency was based on the original sample weights.  3.3.3.2. Recovery ofpaclitaxel from microspheres T o determine the amount of paclitaxel recovered from the above procedure, 5.0 m L of a paclitaxel standard solution was transferred to a test tube. The A C N was allowed to evaporate completely. Approximately 2 m g o f control (no drug) microspheres, accurately weighed, were placed into the tube and dissolved in 1 m L D C M and the resulting solution was transferred to a 100-mL volumetric flask. Paclitaxel was analyzed as described above. The amount o f paclitaxel recovered from the procedure was calculated.  3.3.4. Paclitaxel release studies 3.3.4.1. Paclitaxel standard curves used for HPLC analysis Paclitaxel standard solutions were analyzed before measuring each set o f samples. Paclitaxel standard solutions with concentrations ranging from 0.1 to 50 jag/mL were prepared by diluting a stock paclitaxel solution which was obtained by dissolving about 10 mg o f paclitaxel (accurately weighed) in acetonitrile in a 200-mL volumetric flask.  43  Until use, the solution was stored in a 4 ° C refrigerator. The standard curve was obtained by plotting the peak areas against the paclitaxel concentrations.  3.3.4.2. Determination of chemical stability of paclitaxel in acetonitrile, PBSalbumin, distilled water and methanol Paclitaxel standard solutions were prepared as before and analyzed at weeks 1, 2, 3, and 4, followed by once every month up to 6 months. Paclitaxel standard solution (0.1 m L ) was transferred into a 2-mL H P L C vial, the A C N evaporated completely in the fumehood and 1.0 m L o f methanol, distilled water or PBS-albumin was added to each tube. A l l vials were vortexed for 2 minutes. A t each time interval (t = 0 and every 24 hours up to 7 days), the methanol sample was injected directly into the H P L C . The other samples were prepared by adding 0.5 m L of D C M to extract the paclitaxel. Samples were vortexed for 1 minute, centrifuged at lOOOxg for 10 minutes and the supernatant discarded. The D C M solution was dried completely under a stream o f nitrogen. A C N : H 0 (50:50) (1.0 m L ) was added into each tube. Samples were vortexed for 1 2  minute and analyzed by H P L C as described previously.  3.3.4.3. Determination of in vitro release profiles ofpaclitaxel from microspheres Microspheres (2.5 mg, accurately weighed) were placed into a 15-mL tube and 13 m L of 0.01 M PBS-albumin were added ( N = 4). The tubes were tumbled in the incubator at constant temperature ( 3 7 ° C ) . A t given time intervals, the tubes were centrifuged (550xg, 5 minutes) and 12 m L of supernatant solution were withdrawn. The microspheres were resuspended in 12 m L of PBS-albumin replacement buffer and placed back in the incubator. T o the supernatants was added 1.0 m L D C M . The tubes were vigorously shaken, allowed to settle for 15 minutes and centrifuged at lOOOxg for 10 minutes.  44  Paclitaxel was extracted into the D C M and the supernatant was withdrawn. The D C M was evaporated to dryness under a stream of nitrogen in the heating block ( 4 5 ° C ) . The tubes were reconstituted with 1 m L 50% acetonitrile in distilled water. The solutions were transferred to 1.5-mL Eppendorf tubes and centrifuged at 8000xg for 10 minutes. The samples were analyzed for paclitaxel using H P L C with a mobile phase of water:methanol:ACN (37:5:58) at a flow rate o f 1.0 m L / m i n and U V detection at 232 nm (Burt etal, 1995). The cumulative amounts of paclitaxel released were plotted as a function of time in order to obtain the paclitaxel in vitro release profiles.  3.3.4.4.  Recovery ofpaclitaxel from PBS-albumin Paclitaxel standard solution (500 u.L) was added to a tube. A C N was allowed to  evaporate completely in a fume-hood at room temperature. PBS-albumin (12 m L ) was added into each tube and vortexed for one minute. D C M (1.0 m L ) was added to each tube, vigorously shaken for 10 seconds and then allowed to settle for 15 minutes. The tubes were centrifuged at lOOOxg for 10 minutes and the supernatant removed. The samples were dried under a stream of nitrogen gas at 4 5 ° C for 30 minutes, reconstituted with 1 m L of A C N : H 0 (50:50) and vortexed for 1 minute. The solutions were 2  transferred to 1.5-mL Eppendorf tubes and centrifuged at 8000xg for 10 minutes. The supematants were then transferred to H P L C vials and analyzed by H P L C .  3.3.5. Microsphere degradation studies P L G microspheres (10 mg) with or without paclitaxel loading were placed in a 50-mL tube filled with 50 m L of 0.01 M phosphate buffered saline (PBS) containing albumin (0.4g/L) and tumbled in the incubator at 3 7 ° C . The microspheres were sampled  45  at different time intervals. The buffer was changed either every day for paclitaxel-loaded samples or alternate days for control samples. The microsphere samples were washed with distilled water three times and dried in fumehood and then in the desiccator at room temperature. The molecular weights of the degraded P L G microspheres were determined by G P C . The surface morphologies of the degraded P L G microspheres were observed and recorded using S E M . The glass transition temperature (T ) of each sample was also determined using D S C . g  3.3.6. Co-60 irradiation of PLG microspheres and paclitaxel Pure paclitaxel and the microspheres with or without paclitaxel loading were accurately weighed (100 ± 0.02 mg) and transferred into individual 5.0 m L vials with Teflon® caps. The vials were labeled and shipped to Chiron (US) for Co-60 irradiation. A l l samples were irradiated with a dose of 2.5 Mrad over 350 minutes. Samples were surrounded with blue ice ( 2 ° C ) to avoid heat generation during irradiation. The irradiated samples were characterized using D S C , G P C and H P L C .  3.3.7. Statistical treatment of data Data collected by measurement of several samples from different batches of material were presented as "average ± standard deviation". For A N O V A tests and t-tests, the level of significance was a p-value of less than 0.05 and the hypothesized difference between populations was zero. For particles size distributions of the microspheres, two 60-seconds scans were performed on each sample and data were summarized as the mean ± standard deviation of the particle size distribution.  46  For in vitro release studies, measurements were made on four samples. Values of the average and standard deviation were calculated from cumulative amounts  of  paclitaxel released from each of the four samples. In figures representing these data, the average values are plotted against time and the error bars represent the standard deviations. For all other data collected by repeated measurement of samples from a single batch material, the average was calculated. The relative standard deviation (RSD) of this type of average was used to calculate the precision in the measurements.  47  4.  RESULTS  4.1.  Paclitaxel and 7-epi-taxol Paclitaxel was used as received from Hauser and stored at - 4 ° C until required.  Paclitaxel has two epi-isomers. Paclitaxel has an alpha hydroxy group ( a - O H ) at C-7 while 7-epitaxol has a beta hydroxy group ((3-OH) at C-7. Scanning electron micrographs of paclitaxel and 7-epitaxol revealed that the crystals exhibited a needle shaped crystal habit for both crystals (Figure 9).  4.2.  Physical characterization of P L G polymers  4.2.1. The molar compositions of PLG polymers Polymers were synthesized by B P I in compliance with the F D A guidelines for current good manufacturing practices (cGMP). P L G polymer with a targeted inherent viscosity  (I.V.)  o f greater  than  0.50 were within ± 1 0 % o f its lactide:glycolide  composition. Physical data obtained from BPI indicated that P L G either had different inherent viscosities (I.V.= 0.58-1.06) for the same lactide:glycolide (50:50) composition group of polymers or had different lactide:glycolide compositions for the similar inherent viscosity (I.V.= 0.55-0.60) group o f polymers. P L G polymers contained < 100 ppm o f residual S n  + 2  and < 1%> of residual monomer.  The molar compositions o f the above polymers were determined by H ' - N M R spectroscopy. Representative results are shown in Figure 10. The molar compositions o f the polymers (expressed as a percentage) were calculated from the peak areas o f methylene, methyl and methine proton signals (Amecke et al., 1995). Table 1 compares the L A : G A molar compositions o f the P L G polymers provided by the manufacturer and determined by H ^ N M R . The results o f N M R analysis are in good agreement with the  48  targeted compositions provided by the manufacturer with the exception of two samples of PLG50:50.  4.2.2. Standard GPC curve and molecular weights ofpolymers The elution profiles for polystyrene standards are shown in Figure 11. The calibration curve was obtained by plotting the logarithm of the molecular weights of the polystyrene standards against the peak retention time (Figure 12). Results indicate that the PLgel 10  4  column can be used to measure the polymers with a molecular weight  between 4,000 and 170,600 g/mole.  49  Table 1  L A : G A ratios of the P L G polymers determined using H 1  NMR P L G Polymers L A : G A ratio  ,  b  100:0  J-  85:15 ^  75:25 65:35  '  a  J  j p '  >_1  - '-Mole%  I.V.  Lactide  0.60  IOO+'O.O  0.56  85.2±0.4  « 0.55" 0.55  • 50;50 -  |'  Z-values  ;.  Lactide  Glycolide  0±0.0  N/A  N A  14.8 ± 0 . 4  0.29  Glycolide  3  I  a  3  24.4 ± 0 . 4  d  ().69  I  J  65.7 ± 0 . 6  34.3 ± 0.3  ~  1.35  d  53.7 ± 0 . 7 , . .  46.3 ± 0.5  , -•3.05  c  50:50  0.78  51.0 ± 0 . 4  49.0 ± 0 . 5  1.44  50:50  0."4  50.9 ± 0.4  49.1 ± 0 . 5  i .30  50:50  0.58  52.8 ± 0 . 4  47.2 ± 0.5  4.04  0.29 0.87 0.7  r  1.15  d  d  c  d  1  d  1 ."()4  l!  3.23  c  Molar percents of lactide and glycolide were calculated from the peak area of methine (-CH), methylene ( - C H 2 ) and methyl (-CH ) protons (Amecke et 3  al.,  1995). b  Molar percents of lactide and glycolide provided by manufacturer.  c  Z-values are greater than 1.96 (two-tailed) indicating L A and G A values are significantly different from the theoretical values,  d  Z-values are lower than 1.96 (two-tailed) indicating L A and G A values are not significantly different from the theoretical values.  50  A.  Figure 9  Paclitaxel  S E M micrographs of A : paclitaxel crystal (magnification:  5.0kx at l O k V , 25mA) and B: 7-epitaxol crystal (magnification: 5.0kx at lOkV, 25mA).  CN  IT)  CO  oo 00  cd o  CD  U  O in o in  o  h-l  PH  CO m  a u  I  >n  51  00  J  IT)  o CN  s  a i  CD  ~>^-  V  in  CN  in  O  aA u  h-l  r  0H  i  o  a u5 | • i-H  >  00  u  1  o CU  o o  5h-l  PH  <  Figure 11  G P C elution profiles of polystyrene standards with molecular  weights ranging from 2,000 to 100,000 g/mole at 4 0 ° C (Solvent used: chloroform. Solvent flow rate: 1.0 m L / m i n . Injection volume: 20u.L).  53  LOG(M.W.) = 8.551 - 0.567t, r = 0.998 2  ^  55  5.0  5.5  6.0  6.5  7.0  7.5  8.0  8.5  9.0  t (min.) Figure 12  GPC  standard  curve  for  polystyrene  standards  with  molecular weights of 4, 9, 17.5, 30, 50, 100 and 170 kg/mole on a PLgel column with a nominal pore size of 10  4  A  at 4 0 ° C (Solvent used:  chloroform. Solvent flow rate: 1.0 mL/min. Injection volume: 20 p L ) .  54  4.3.  Size distribution of the microspheres The size distributions of dried microspheres were determined by laser scattering.  The small size microspheres showed 98% (by volume) o f microspheres falling into the range o f l-20um as shown in Figure 13 A . The large size microspheres had a size range between 20-100um, as shown in Figure 13B, with over 95% (by volume)  o f the  microspheres falling into this range.  4.4.  Surface morphologies of microspheres The  surface morphologies o f microspheres were revealed  with  SEM.  An  extensive pore network (pinholes) was found on the surface o f 20% paclitaxel-loaded PLG100:0 (I.V.= 0.60) microspheres (Figure 14A). This surface morphology was not seen on the 10% paclitaxel-loaded (Figure  15A) or control PLG100:0 (I.V.= 0.60)  microspheres (Figure 16A). The control, 10%> and 20% paclitaxel-loaded  PLG85:15  (I.V.= 0.56), PLG75:25 (I.V.= 0.55) and PLG65:35 (I.V.= 0.55) microspheres had very smooth surface morphologies as shown in Figures 14B, C , D and 15B, C , D as well as in 16B, C , D . The S E M micrographs of control, 10% and 20% paclitaxel-loaded  PLG50:50  (I.V.= 1.06), PLG50:50 (I.V.= 0.78,) PLG50:50 (I.V.= 0.74), and PLG50:50 (I.V.= 0.58) microspheres are shown in Figures 17-19. N o differences in the surface morphologies were observed among these microspheres. The S E M micrographs o f PLG50:50 (I.V.= 1.06)  microspheres with 1 to 5% paclitaxel loading also revealed that their surface  morphologies were smooth and there were no differences due to different paclitaxel loading (Figure 20).  55  Differential Volume  0.6  1  T  4  1000  6  10 20 40 60 Particle Diameter (um) L O 0.974 um U O 19.65 um (99.49%)  Size distribution of small (1-20LUTI) microspheres (mean = 3.4 ± 1.5pm) Differential Volume  4  10 20 40 60 100 Particle Diameter (um) LC= 19.65 um UC= 101.2 um (95.65%)  B.  6  200  400  1000  Size distribution of large (20-100urn) microspheres (mean = 65 ± 21pm)  Figure 13 Representative size distributions of A : small and B: large sizes of microspheres determined by the laser scattering particle size analyzer. The size distributions of the microspheres were the averages of the two 60-seconds scans. Microspheres were resuspended in polysorbate 80 solution.  A . PLG100:0 (I.V - 0.60) microspheres  B. PLG85:15 (I.V.= 0.56) microspheres  C . PLG75:25 (I.V.= 0.55) microspheres  D. PLG65:35 (I.V.= 0.55) microspheres  Figure 14 microspheres  S E M micrographs [A: PLG100:0,  of  20%  paclitaxel-loaded  PLG  20-100um ( l k x at 5 k V , 25mA); B:  PLG85:15, l-20um (lOkx at 6kV, 25mA); C : PLG75:25, l-20um (4kx at 6kV, 25mA); and D: PLG65:35, l-20um ( l k x at 5kV, 25mA)].  57  A . PLG100:0 (I.V.= 0.60) microspheres  B . PLG85:15 (I.V.= 0.56) microspheres  C. PLG75:25 (I.V.= 0.55) microspheres  D. PLG65:35 (I.V.= 0.55) microspheres  Figure 15  S E M micrographs  microspheres  [A: PLG100:0,  of  10%  20-100pm  paclitaxel-loaded  PLG  (2kx at 5 k V , 25mA);  B:  PLG85:15, l-20u,m (lOkx at 6kV, 25mA); C : PLG75:25, l-20u.m (600x at 5kV, 25mA); and D: PLG65:35, l-20um (lOkx at 6kV, 25mA)].  Figure 16  SEM  micrographs  of  control  PLG  microspheres  [A:  PLG100:0, 20-100um (1.5kx at 5kV, 25mA); B: PLG85:15, 1-20p.nl (6kx at 5kV, 25mA); C : PLG75:25, l-20um (3kx  at 5kV, 25mA); and D:  PLG65:35, l-20um (2kx at 7kV, 25mA)].  59  A. PLG50:50 (I.V.= 0.58) microspheres  B . PLG50:50 (I.V.= 0.74) microspheres  C. PLG50:50 (I.V.= 0.78) microspheres  D. PLG50:50 (I.V.= 1.06) microspheres  Figure 17 S E M micrographs of 20% paclitaxel-loaded PLG50:50 microspheres [A: I.V.= 0.58, l-20pm (lkx at 5kV, 25mA); B : I.V.= 0.74, l-20um (lOkx at 5kV, 25mA); C: I.V.= 0.78, l-20u.m (2kx at 5kV, 25mA); and D: I.V.= 1.06, l-20um (3kx at 5kV, 25mA)].  60  A . PLG50:50 (I.V.= 0.58) microspheres  B . PLG50:50 (I.V.= 0.74) microspheres  C . PLG50:50 (I.V.= 0.78) microspheres  D. PLG50:50 (I.V.= 1.06) microspheres  Figure 18  S E M micrographs o f 10% paclitaxel-loaded  PLG50:50  microspheres [A: I.V.= 0.58, l-20um ( l k x at 5kV, 25mA); B : I.V.= 0.74, l-20um (5kx at 5kV, 25mA); C : I.V.= 0.78, l-20um (5kx at 5kV, 25mA); and D: I.V.= 1.06, l-20um (4kx at 5kV, 25mA)].  61  A . PLG50:50 (I.V.= 0.58) microspheres  B . PLG50:50 (I.V.= 0.74) microspheres  C. PLG50:50 (I.V.= 0.78) microspheres  D. PLG50:50 (I.V.= 1.06) microspheres  Figure 19  S E M micrographs of control PLG50:50 microspheres [A:  L V = 0.58, l-20um (400x at 5kV, 25mA); B : I.V = 0.74, l-20pm (2kx at 8kV, 25mA); C : I.V.= 0.78, l-20um (2kx at 5 k V , 25mA); and D: I.V.= 1.06, l-20um (5kx at 5kV, 25mA)].  62  A . 1% paclitaxel-loaded  Figure 20  B. 5% paclitaxel loaded  S E M micrographs of paclitaxel-loaded PLG50:50 (I.V.=  1.06, l-20u.m) microspheres [A: 1% paclitaxel (4kx at 5kV, 25mA); B : 5% paclitaxel, (lkx  at 5kV, 25mA); C : 10% paclitaxel, (4kx at 5kV,  25mA); and D: 20% paclitaxel, (3kx at 5kV, 25mA)].  63  4.5.  Characterization of microspheres  4.5.1. X-ray diffraction patterns of the microspheres Figure 21 shows the X-ray diffraction pattern o f paclitaxel as received from Hauser. Paclitaxel gave three strong diffraction peaks between 5-15° (29). Figure 22 shows the X-ray diffraction patterns of 20% paclitaxel-loaded and control P L G 100:0 (I.V.= 0.60, 20-100pm) microspheres. A typical amorphous halo pattern was observed for all control and paclitaxel-loaded microspheres. X-ray data for all microspheres are given in Table 2.  o o o  1 1 1 1  1  I'I'I'I  50. Diffraction angle ( ° 2 9 ) Figure 21  X-ray powder diffraction pattern of paclitaxel. Paclitaxel  sample as received was scanned at 2 7 m i n at 20mA and 40kV at room temperature ( 2 5 ° C ) .  64  Figure 22  X-ray  diffraction  paclitaxel-loaded PLG100:0  patterns  of A :  control  and B :  20%  microspheres. Microsphere samples were  scanned at 2 ° / m i n at 20mA and 40kV at room temperature ( 2 5 ° C ) .  65  Table 2  X-ray diffraction results for control and paclitaxel-loaded  P L G microspheres. Microsphere samples were scanned at 2 ° / m i n at 20mA and 40kV at room temperature ( 2 5 ° C ) Polymer  i  Paclitaxel Loading  0  Amorphous  0  Amorphous  0  Amorphous  0  Amorphous  0  Amorphous  0  0%  Amorphous  0  o%  Amorphous  0  • 20%  Amorphous  0  20%  Amorphous  0  Amorphous  0  Amorphous  0  Amorphous  0  Amorphous  0  Amorphous  0  Amorphous  0  Amorphous  0  Amorphous  0  - Amorphous  0  Amorphous  0  a  "0.56  h  PLG75:25  0.55  b  PLG65:35  0.55  b  PLG50:50  1.06  b  PLG50:50  ; . 0.78  b  0% .  PLG50:50  0.74  b  i  1M.G85:I5  PLG5():50  i— "6.5 8  PLG100:0  r  0.56  PLG85:15 IM.G75:25  r  PLG65:35 t  PLG50:50  :[ .....  :  .  ().55  b  -  O/o  li •:•  •L  (  1  . . .  ™ .... *  0.74  PLG50:50  0.58  PLG50:5"  l".06  20% . T ^ T  2 0 O / o  • • „ ,; ,,„, , * F  20%  a  11  20%  b  1.06  b  .  1.06  :  '  TTT7  b  ,;  ,;  '  "  :  JT"  . . .  b  ;  „ „..  L  20%"  h  PLG5():50  ;•  20%  LOG" 0.78  •  0%  b  0.78  PLG50:50  - |  a  PLG50:50  PLG50:5()  0%  b  0.55  PLG5():5i"  0%  h  o.6o  - X-ray Results •  .  Amorphous  0.60  PLG100:0  ['-' i  I.V.  i%  a  Microsphere size = 20- 100pm.  b c  Microsphere size = l-20um. A typical halo pattern as shown in Figure 22 was observed.  - .'•  4.5.2. Glass transition temperatures of control and paclitaxel-loaded microspheres The D S C thermograms of paclitaxel as well as control and 20% paclitaxel-loaded P L G microspheres are given in Figures 23 and 24. The paclitaxel sample had a melting peak  at 2 2 4 ° C (Figure  23A). Exothermic  peaks  at 2 3 0 - 2 4 0 ° C were due to the  decomposition o f paclitaxel at high temperatures. However, no paclitaxel melting peaks were found in any o f the paclitaxel-loaded P L G microspheres (Figure 24A). The exothermic peaks at about 2 5 0 - 2 6 0 ° C were observed in the paclitaxel-loaded P L G microsphere samples, similar to the paclitaxel sample. The D S C thermogram o f the control P L G microspheres did not show any exothermic peak after 2 3 0 ° C (Figure 24B), but the decomposition of the P L G polymer was found after 2 9 6 ° C . The glass transition temperatures  o f the microsphere samples are shown as  endothermic peaks at about 5 1 ° C in Figures 23 and 24 and are summarized in Table 3. The glass transition temperatures increased with an increase in the paclitaxel loading in the microspheres with either different L A : G A compositions or the same 50:50 L A : G A composition but different molecular weights. The p-value  from two-tailed tests indicated  that increases in the glass transition temperatures were significant for all 20%> paclitaxelloaded microspheres regardless o f the sizes and L A : G A  compositions o f the P L G  microspheres.  4.5.3. Fourier transform infra-red spectroscopy (FTIR) Figure 25 shows the F T I R spectrum o f paclitaxel and the F T I R spectrum o f control PLG50:50 (I.V.= 0.74, l-20um) microspheres. Figure 26 shows the F T I R spectra of a physical mixture of paclitaxel/control PLG50:50 (I.V.= 0.74, l-20pm) microspheres and  20% paclitaxel-loaded  PLG50:50  (I.V.=  0.74, l-20um)  microspheres. The  67  wavenumber o f paclitaxel C = 0 in-plane (5C=0) and out-of-plane (yC=0) deformation was 710 cm" while the wavenumbers o f C = 0 in-plane (5C=0) and out-of-plane (yC=0) 1  deformation from paclitaxel-loaded microspheres increased significantly (about 721-730 cm" ). The p-values 1  for the two-tailed student t-tests revealed that the wavenumbers of  C = 0 in-plane (8C=0) and out-of-plane (yC=0) deformation for paclitaxel-loaded P L G microspheres were increased significantly compared to control P L G microspheres (Table 4).  68  as ca'  5  •.f  "tf  - CN  ,©•.1 o  o •! o  •Si co  "3  vo  O  © ©"  o o  O  V  C3  IIP —  o  ©  ©  .©'  CN O  o  g  CD X  CD t-i  03  CD  a  CD  CO  2 CN  o  03 O H  O'  0>  + 1  © +1  Q \  t--  in" in.  rr! in  DC  CN  o + 1  in  , + 1  •  + 1  CD  CN  VO  in  ©  00  ©  co  — i  O  03 O H  +i  + 1  t  00  'm  1  vo  CD  CD  "3  o" + 1  , o  ON  O  -'-DO  CD  I H, '  © +1 in CN in  + 1  . + 1  + 1  in  r»n  *—H  m  in  03  CN  1  '—  +1  + 1  ,  i>n-  ^f" m  >n  co o  o  .  O .  -1 s °  ;•• CD -  « s  O  c o  o o  <3 O H  CD  X cd  C M  CO CO  03 O H  +1 CN rf in  O  I  •  CD  co .3  ©"  ©  + 1  + 1  rN  ^o  oo  :  o  ©'  ©  in  Mil  >n  in  ro in  + 1  + 1  c ©  o  ©  CN  o  o  CN I  GO  O  CD T3  CD  CD  T3  O  EllilSlJl  CN  —  in  T3  & VI  to  "  ^  C  3 S  O CN  rN  « o  rT  o  A O  in  m •m  ;o"  -—-  in '©'. ,<n  .o  •h-l  o  in © in  m ©  o  a  PH  CPH  h-l  in oo  m CN in  o  o  PH  p-  hJ  h-l-  m .in  "i  o  -1ro in  , O PH  J3  ^—»  co  + 1  O H JH  II  co  7^ CJ  hH  CD  CD CD  CO  CO  (O  "2 .3 03  03 s  hH  o^  ^3 O  CD  -i  •o  in  © o3 S o  CD  CO  03  o C ^H S }> o"  +H  -O 00  'I  + 1  CN  CN  S3 O  .ti  ">  ~  ©'  03  CD  ^  o  CD  >3  i"  S 'I 3  izo  y  —  cr  L CD >  .u  1 ^  a "5b 00  O H  P 'rj  03  n.  g X  CD  SOr 1 in  i  *-I  CO  +1'  o  M  co  03 co  -°  ,:r-H  •  l-l  3  CD  CD  0  SIS  ..1 2  CD  O  >  © ?3  Sb  § '3  CD CD  ^ . ,  o  o a  EH  o •  O  1  ^3 _  -  o  -1  CD  •§ 1  PH  r  co 03  .a  o  CN  m  h-H  CO  CD  03  ' o  CO  CD  CD  CO  CD *H  X  C/3  CO  CD  o  ! | >s 03  o  CD  e^ hS  2 " n o  ^  a  0.5  o X  236.57°C  w  o.o A  242.96°C  00  -0.5 H  o  E  -1.0  -1.5 223.84°C  o -2.0  50  100  150  200  250  300  Temperature ( ° C ) A : D S C thermogram of paclitaxel 0.5  o X  0.0  00  V o  -0.5  E  \1  217.69°C  51.66°C  -1.0  o  w -1.5  50  100  150  200  250  300  Temperature ( ° C ) B: D S C thermogram of physical mixture of paclitaxel and control PLG50:50 (I.V.= 1.06, l-20um) microspheres (1:4) Figure 23 mixture  D S C thermograms  o f paclitaxel/PLG50:50  o f A : paclitaxel (I.V.=  and B : a physical  1.06, l-20pm)  microspheres  (paclitaxel: microsphere ratio is 1:4). Samples were scanned at a heating rate o f 10 ° C / m i n with a N 2 purging rate of 40 m L / m i n .  1.5 1.0 (50  256.90°C  0.5 0  o  -0.5  CO CD  -1.0  i  -1.5  O  57.49°C  -2.0  C -2.5 0  50  100  150  200  250  300  Temperature (°C) A : DSC thermogram of 20% paclitaxel-loaded PLG85:15 (I.V.= 0.56, l-20pm) microspheres  Temperature (°C) B : DSC thermogram of control PLG65:35 (I.V .= 0.55, 120u.m) microspheres Figure 24 DSC thermograms of A : 20% paclitaxel-loaded PLG85:15 (I.V.= 0.56, l-20um) and B : control PLG65:35 (I.V.= 0.55, l-20pm) microspheres. Samples were scanned at a heating rate of 10 °C/min with a N purging rate of 40 mL/min. 2  Table 4 Wavenumbers of C=0 double bond in-plane (8C=0) / outof-plane (yC=0) deformation absorption of the P L G microspheres (+ paclitaxel). Samples were run at a resolution of 2 cm" . Pokmcrs  I.V.  Wavenumber of control P L G microspheres (cm )  \\'a\ cnumber of 20" o paclitaxel-loaded PLCi microspheres (cm"')  1  p-valuc"  h  0.009  PLG 100:0  0.60  707 + 4  730 ± 8  PLG85:15  0.56  708 = 2  "27 + 3  PLG75:25  0.55  709 ± 2  727 ± 4  PLCK#3T^  0.55  70S + 2  726,+ 3  o.oof  PLG50:50  1.06  708 = 1  724 ± 4  0.002  PLG50:50  0.78  710 ± 2  726 + 7  0.02  PI ( o n  SiI  0.74  72\ - 4  0.01  PLG50:50  0.58  730 ± 6  0.003  a  Value  obtained  708 ± 2 by  two-tailed  student  t-test  • 0.001  '  assuming  0.002  equal  variances.  Wavenumbers of paclitaxel-loaded microspheres are significantly higher than control microspheres when p < 0.05. b  Data are the average of three samples, ± standard deviation.  72  %Transmittance  102.00  87.60 -i  73.20  58.80  44.40 H  30.00 3200  3900  A.  2500  1800  1  — f  .  1100  .  400 Wavenumber (CM-1)  1  r  F T I R spectrum of paclitaxel as received  % Transmittance  48.00-1  29.00-1  Wavenumber (CM-1)  B.  F T I R spectrum of control PLG50:50 (I.V.= 0.74, 120pm) microspheres  Figure 25  Representative F T I R spectra of A : paclitaxel and B : control  PLG50:50 (I.V.= 0.74, l-20pm) microspheres. Samples were run at a resolution of 2 cm" . 1  % Transmittance  -|  3900.  .  A.  •  1  ,  3200.  •  1  .  1 2500.  1  •  .  1 1800.  .  i  .  1 1 1100.  •  1400.  1  Wavenumber (CtVM)  F I T R spectrum of control PLG50:50 (I.V .= 0.74, 120pm) microspheres & paclitaxel mixture  % Transmittance  3900.  B.  3200.  2500.  1800.  1100.  400. W a v e n u m b e r (CM-1)  F I T R spectrum of 20% paclitaxel loaded PLG50:50 (I.V.= 0.74, 1-20pm) microspheres  Figure 26  Representative F T I R spectra o f A : a physical mixture of  control PLG50:50 (I.V.= 0.74, l-20pm) microspheres/paclitaxel and B : 20% paclitaxel-loaded  PLG50:50 (I.V.= 0.74, l-20um) microspheres.  Samples were run at a resolution of 2 cm" . 1  74  4.5.4. Total content and encapsulation efficiency ofpaclitaxel in microspheres Paclitaxel recoveries during extraction, expressed as a percentage, are shown in Table 5. A high percentage of paclitaxel ( « 99%) can be recovered from this process. Table 6 summarizes the encapsulation efficiencies o f different paclitaxel-loaded PLG  microspheres. The microsphere  fabrication  process had high  encapsulation  efficiency for paclitaxel. Most microspheres had paclitaxel encapsulation efficiency o f greater than 90% except for 2 0 % paclitaxel-loaded PLG100:0 (I.V.= 0.60, 20-100um) microspheres. The actual loading of paclitaxel in the microspheres is listed in Table 7.  4.5.5. Stability of paclitaxel in acetonitrile, PBS-albumin, distilled water and methanol Paclitaxel was very stable in A C N with less than 0.5%> o f paclitaxel degrading over one month (at 3 7 ° C ) or over six months (at 4 ° C ) . Paclitaxel in methanol and distilled water were relatively stable within 72 hours, with less than 1% o f paclitaxel degrading (at 3 7 ° C ) . Paclitaxel was less stable in PBS-albumin. About 2% o f paclitaxel degraded in 24 hours ( 3 7 ° C ) . However, no 7-epitaxol peak was found in the H P L C chromatogram during release studies.  75  Table 5 Percentages of paclitaxel recovery during extraction from control P L G microspheres/paclitaxel mixtures Polymers  I.V.  % of paclitaxel recovery  PLG100:0  0.60  98.8 ± 1.0  PLG85.15  0.56  99.5 ± O.S"  PLG75:25  0.55  98.3±0.8  PLGG5.35  0.55  y7.5 ± 1.2''  PLG50:50  1.06  99.0 ± 0.9  ~PLG5,0:50"  a  a  97.9 ± 0.9"  PLG50:50 PLG50:50  a  0.74 0.5S  99.0±0.8 ; .1  a  99.8+ 1.0  a  1  Average  98.7 ± 1 . T  a  Data are the averages of three samples (N = 3), ± standard deviation,  b  Average of paclitaxel recovery from all P L G polymer  samples during  extraction, ± standard deviation.  76  Table 6  Paclitaxel encapsulation efficiency in P L G microspheres  with different loading, sizes and L A : G A compositions Polymers •  I.V.  j  ' Microspheres  J  I-  :  :  ! :  j. sizes (u.m) • 1 -20  ; PLG50:501 ' 1.06--"! 1  Encapsulation efficiency (%) o f paclitaxel in P L G  f l l p | £ ;.M%i •''" ':vi n!i 'i t H ii WM1 Si! 1.  •  microspheres i% paclitaxel  5% b  Um • l.0  J  paclitaxel  20%  10% paclitaxel  b  i 100± rj^2 ; " 9 3 . 4 ± 6 . 8 :  :  a  A  paclitaxel  3  95.S ± 0.o  J  >..*  PLG50:50  0.78  2 0 - 100  94.7 ± 0.6  PLG5():50  0.78  1 -20  90.3 + 0.9 ' 1  90.71  PLG50:50  0.74  91.8±0.7  a  92.1  PLG5():50  '" 0.58 "  PLG10U:U  0.60  a  93.4+ 1.0  a  l.l'  1  10.6  a  p*|^^Bfe|^liMi:| P !?i i fl ^ fill !'i i'i W. hi f! f  1-20 ;  92.6 ± 0.9'  95.7 ± 1.0  95.9±0.8  86.5 ± 0.6  a  94.0 ± 0.4  J  1  20 - 100 !  1-20  a  ; 99.3 ±0.<Y' :  [PLG85:1S I r<>.5o PLG75:25 0.55  1-20  93.7±0.6  rpLG65-35" W55.':  1 -20  92.8 ± O  a  b  a  1 p i ! Iffitsf iFlil't fy Sfffttfl I! (1 !t| I i i f  x  a  96.8 ± 0.5  a  96 5 ± ().5  J  Data are the averages of three samples, ± standard deviation. Encapsulation efficiency (%) = 100 x [(paclitaxel content) / (paclitaxel + polymer content)] x [(1/ (initial paclitaxel loading)]  b  Numbers are the theoretical  calculation based on the weights  of P L G  polymers and paclitaxel.  77  Table 7  Actual loading of paclitaxel in P L G microspheres with  different l o a d i n g , sizes and L A : G A compositions • Polymers  IA'.  Microspheres  !  Actual paclitaxel loading in P L G microspheres  si/os (pm)  1.06  PLG50:50  0.78  PLG50:5()  0.78  20  PLG50:50  0.74  1 - 20  fPLG50  • 0.58 1-• •  PLG100:0  0.60  0.55  PI i i i . 5  0.55  "  •'  a  5.0  a  ~|  20%  Paclitaxel  15  2 0 - 100  r  b  a  19.2'  9.5  a  18.7  a  18.1  a  9  •'  18.4  a  T,  illlHHj  i - 20  1HI ^ i l l l l l l — La.i>*i L J • -  -  9.6  a  9.9  J  9.4  a  b  1  -° a  9.2  2 0 - 100 20  Paclitaxel  9.3  i l i i R .20  , PI C i S S : l 5 ; 0.50 PLG75:25  5% Paclitaxel  1.0  20  PLG50.50  10%  i% Paclitaxel  19. P* , > 17.3  f  a  a  18.9'  !  19.4  a  19..V  J  a  Actual loading of paclitaxel in the microspheres (N = 3).  b  Numbers are the theoretical loading calculated based on the weights of P L G polymers and paclitaxel.  78  4.5.6. Release studies of paclitaxel-loaded microspheres 4.5.6.1. Statistical treatment of release study data The two-tailed t-tests were used to compare the cumulative amounts of paclitaxel released on each pair of release curves involved for comparison on day 1, 2, 5, 10, 15, 20, 25 and 30 i f applicable. When the  p-value was < 0.05, the difference between the  amounts o f paclitaxel released would be considered significant. The slopes (rate o f paclitaxel released) of linear regressions on each of three segments (segment 1: day 1 to day 2; segment 2: day 5 to day 10 and segment 3: day 15 to 25) from the same release curves, or on the same segments of different cumulative amounts of paclitaxel released curves were also compared.  4.5.6.2. Paclitaxel release from PLG microspheres Figure 27 shows the paclitaxel release profiles from PLG50:50 (I.V.= 1.06, 120pm) microspheres with paclitaxel loading from 1 to 20%. A burst phase of paclitaxel release was observed from the microspheres during the first 2 days; then followed by a relatively steady release of paclitaxel from microspheres from day 4 to day 28. The 20% paclitaxel-loaded microspheres showed the fastest release rate (amounts o f paclitaxel release per day) than any other microspheres. Two-tailed t-tests on the cumulative amounts o f paclitaxel released at about day 5, 10, 15, 20, 25 and 30 indicated that the differences among the amounts of paclitaxel released for the 1, 5, 10 and 20% paclitaxelloaded PLG50:50 (I.V.= 1.06, l-20um) microspheres were significant  (p-values < 0.05).  Figure 28 shows the release profiles of paclitaxel from PLG100:0 (I.V.=0.60, 20100pm) microspheres with 10 and 20%> paclitaxel loading. There was a very large burst  79  phase for the 20% paclitaxel-loaded microspheres, which was followed by a period o f slow release to approximately day 6, then another phase o f fast release. Figures 29 and 30 show the release profiles o f paclitaxel from PLG85:15, PLG75:25 and PLG65:35 microspheres with the same size ranges (l-20pm), similar inherent viscosities (I.V.= 0.55 - 0.56) and 10 and 20% paclitaxel loading. A burst phase of paclitaxel release was observed during the first 2 days; this was then followed by relatively steady release. The cumulative amounts of paclitaxel-released at day 23 for the PLG75:25 and PLG65:35 microspheres were similar and were significantly higher than the cumulative amounts of paclitaxel-released for the PLG85:15 microspheres (t-tests). Figures 31 and 32 show the release profiles for 10 and 20%> paclitaxel-loaded PLG50:50 microspheres with the same size ranges (l-20pm) and different inherent viscosities (I.V.= 0.74 - 1.06) or molecular weights. Burst phases were observed up to day 2. Paclitaxel release was significantly lower for 10%> paclitaxel-loaded PLG50:50 microspheres with I.V.= 1.06 compared to microspheres either with I.V.= 0.78 or with I.V.=  0.74. For 20%> paclitaxel-loaded microspheres, the cumulative  amounts o f  paclitaxel released were significantly different for all microspheres with different molecular weights. Figure 33 compares the release profiles o f two different sizes (1-20 p m and 20100 urn) o f 10% paclitaxel-loaded PLG50:50 (I.V.=0.78) microspheres. The burst phases of these two sizes o f microspheres were observed during the first two days; this was followed by a period of relatively steady release. The microspheres with smaller size range produced slightly higher burst phase release than microspheres with larger size  80  range. The rates of paclitaxel release from both size ranges of the microspheres were not significantly different (t-tests).  81  200  Time (days) Figure 27  Cumulative amounts of paclitaxel released from PLG50:50  (I.V - 1.06, l-20um) microspheres in PBS-albumin at 3 7 ° C (N = 4).  82  120  1 0 % paditaxel-loaded 2 0 % paclitaxel-loaded  0  10  15  20  25  30  35  Time (days)  Figure 28  Cumulative amounts of paclitaxel released from PLG100:0  (I.V.= 0.60, 20-100pm) microspheres in PBS-albumin at 3 7 ° C (N = 4).  83  120  0  5  10  15  20  25  T i m e (days)  Figure 29  Cumulative amounts of paclitaxel released from 10%  paclitaxel-loaded P L G microspheres (l-20u.m) with different L A : G A compositions in PBS-albumin at 37°C (N = 4).  84  200  Time (days)  Figure 30  Cumulative  amounts of paclitaxel  released from  paclitaxel-loaded P L G microspheres (l-20um) with different compositions in PBS-albumin at 3 7 ° C (N = 4).  20%  LA:GA  140 ~  120-  Time (days)  Figure 31  Cumulative  paclitaxel-loaded  amounts  P L G microspheres  o f paclitaxel  released  from  (l-20u.m) with a 50:50  10%  LA:GA  composition in PBS-albumin at 3 7 ° C (N = 4).  86  250  Time (days)  Figure 32  Cumulative  amounts  of paclitaxel  released from  paclitaxel-loaded P L G microspheres (l-20um) with a 50:50  20%  LA:GA  composition in PBS-albumin at 3 7 ° C (N = 4).  87  Figure 33  Cumulative amounts of paclitaxel released from 10%  paclitaxel-loaded  PLG50:50  (I.V.=  0.78,  l-20um  &  20-100pm)  microspheres in PBS-albumin at 37°C (N = 4).  88  4.6.  In vitro degradation of microspheres in PBS-albumin buffer  4.6.1. Molecular weights of degraded PLG microspheres determined by GPC Figure 34 shows the G P C elution profiles for P L G polymers used in the studies. Table 8 summarizes the molecular weights of control PLG50:50 microspheres of differing inherent viscosities degraded in PBS-albumin at 3 7 ° C . Table 9 provides the molecular  weights  of  control  PLG100:0,  PLG85:15,  PLG75:25  and  PLG65:35  microspheres of similar inherent viscosities (I.V.= 0.55 - 0.56) after incubation in P B S albumin at 3 7 ° C . Table 10 shows the molecular weights of 10%  paclitaxel-loaded  PLG85:15, PLG75:25, PLG65:35 and PLG50:50 microspheres which were incubated in PBS-albumin at 3 7 ° C . Degradation profiles for control PLG50:50 microspheres with differing inherent viscosities (I.V.= 0.58 - 1.06) were obtained by plotting their molecular weights (MGPC) against degradation time in the PBS-albumin (Figure 35). The molecular weights of the PLG50:50 microspheres decreased over a period of 4-6 weeks. The higher the initial molecular weights, the faster the degradation rate of the polymers. Figure 36 shows the degradation  profiles  of  control  PLG100:0,  PLG85:15,  PLG75:25  and  PLG65:35  microspheres with similar inherent viscosities. The molecular weights of these P L G polymers remained quite stable over 4-6 weeks with little or no decrease in molecular weight, compared with P L G polymers with a L A : G A ratio of 50:50. Figures 37 and 38 show the degradation profiles o f control and 10%> paclitaxelloaded PLG75:25 and PLG65:35 (Figure 37) and PLG85:15 and PLG50:50 (Figure 38) microspheres. There was no significant difference between the degradation profiles of the control and 10% paclitaxel-loaded microspheres (p > 0.05).  89  Figure 39 compares the degradation profiles of two different sizes ( l - 2 0 p m and 20-lOOu.m) of control PLG50:50 (I.V.= 0.78) microspheres. The smaller microspheres appeared to  show faster  degradation  over the  first  2 weeks  compared to  larger  microspheres.  90  Pl_Q30:S0 (I.V.=1.08)  A:  G P C elution profiles of P L G polymers with the same L A : G A composition (50:50) but different inherent viscosities.  B:  G P C elution profiles of P L G polymers with different  L A : G A compositions but similar inherent viscosities (I.V. « 0.55).  Figure 34  G P C elution profiles of A : P L G polymers with the same  L A : G A composition (50:50) but different inherent viscosities and B : P L G polymers  with  different  LA:GA  compositions but  similar  inherent  viscosities (I.V. « 0.55) at 4 0 ° C (Solvent: chloroform. Solvent flow rate: 1.0 mL/min).  91  CO  CN  O  as  +1  CN  rf  CO  oo  ro  —  +1  ro IT)  CO  CD  CO  00  CO  co O O  U  oo  o  i3 CD  —  o  i t  S3 CD  CO 03 CD  a £ q  CD CD  43 &  CO O  43 00  .2 "53  CD  •a  CN  o  S3  >  75 3  •i  a  43  43  '53  oo CQ  73i PH  61  : (3  -a  ON  +1  ro  rr  p$pti Os r-  o  +1  fN  ro ©  rr  +1 '  CN  +1  oo rr ro  ro ©  [ CN  in ©  1—H  d  +1  ro  +1  NO  lEGl  T3  00  : ro  CD  q —•  o  d in  1  1 0  ! I> 1 rT  NO ©  ro ro  **  +1 r-©  d O  +1  r-[ON  l-  -  +1  +1  CN  © CN  +1  as +1  o d  m d  +1 X ©  +1  CN  NO  +1  in in  d  d  rr d  o  *—i  NO  oo rf NO  CN  a  ° S3  rf  o d  q  +1  NO  m C""  -t rr  m  CN  NO  +1  T3  CD j>  <J~t  ©  O'" •ro  d CN  ©  +1  t  1111d ri r^  +1  OO  rr d  +1 rf d  +1  d  Os-  +1  ro  q  NO  +1  q  ON  75 CO  03 CO CD S3 CD  T3 03 00  3  O  o  ©  oo  CD  i-  rr  CO  03 T3  03  CD > O  E-1  1  CD  I-I  CD  CD  v-  3  oo  r.l  "ii  2 o  PH  o  ,.0D  -cs  O CD  h-l  ChH O  *©'  CO  73.1  o CD  O in d m  00  l l i i i l  r-i  75 K  l-i  CO CD ><D  "ro  00  03  CD  43  CN  CN  O rr  E3 OH  qin  ©  75  ex  '•3  < hJ  N? J H  75  a  q oo  —  iO  1/-I d  , IT)  —  ri +1  oo d  o in d  in  CD  00  if < a  rr i  o CN  , CD  +1  NO  m in  o o c  d  CN  +1  o in d in  .o CN  r-i  '•a , !O0  CO  ,j  >• i  OO  oo  rf  03  CO  >,  fr>s  ON  cO  13 CN Tt  oo  +1  Tt  VO  ro  b  ON  +1  CO • .  in  1 U  o  o  cO 13 00 CN  >. l-i  CO  00  43  op '53 o  <u  CO  o  ,S:-  a  e i  PQ  O co  43  'C  e > O  o  CD  "3  CD 00 <D  13  CD O  CD  CD  Tt  oo  r-H  CN  i  VO  b  T t  b  +1  oo  o  ^~  j2  +1  +1  in in  CN  Tt  +1  +1  °P*"+I  Tt  ••3/ &.  oo  b  CO  >> cO ,73 ICO  (3 CD  S3  o  "§  00  CN  00 ON  ^ +|  0\  ro  in  +1  +1  b  Irtiliil r i CN." O  •> 13  CO  i-sSSSJ  t  *ti > i  o o  b  oo  id S3 +1  CO  p  p  CN ON  b  +1  +1  CO  +1 VI S3  P co ' >> cO 13 CN  CD S3  <u  in CN  CN  +r  ON  rt  .+1  CO 0 0  oo  •  CN CN  +i  Tt  CD  b  +l  &  1 CO  u  CM  CO  -f  1  co  13  >n  VO  ON  +1  ON  +1  b  CN ON'  vd  +1  ^t  ON  b  +1  O  ^—•  o  CO CO  ON,,  V O  CO  13  <D S3  CN  +1  ON  +1  vd  +1' S  >>  <  O  'o  o o  B  in  ;in  oo  'co  o  43  +-»  CD  in CN  !, ? v  CO  ij  J  in vo  j  o  o  T t  CD 00 CO iCO (D  ''I  in  *—*  b +l  ifHi  co  'o a oo .S3  CN  00  CO  co  ON  +1  r-  -  13  13 cO  co  CO •13  8 CM j> _c o 00 CD PH  ro i OYt  ^ '  S3  to •  8 m o  •£  !  CD  'o  CO  co  _ L  <N  c  "o o  m  1"-  43  43 &  +1  '  »H  l-l  <D )-<  +l  (3  CO (L)  CD  "•3  *  ON.  <L>  (U  -4-* S3  H  '-t CN oo  13  IO s  < 1-1  CN  +1  CO  o  o < o  b  ae  S3  Tt  ro  |  hre  CO  +  ve  Id  00  CO  CO  co  OO  £ 13  CO CO 43  op 'S cO  O O  o  CN  o  o  CN  CN  o CD  m m  CO  m  o  CN  o ,vO'  b  VO in  Tt ON  cd  CO  o  U  Cd OH  00 b  ^H  00  O ^H  r-  O  Td  O CO £3 O  CO O  OH  a  I;S2  id CJ £3  ca  r-  o  T t  co  <u  •12  -a  '53  o  43  60  ON •NO  43  co  cj  (-1 CJ  43 PH CO  8 CJ  o  h-l  CJ  <2  1 PH .£3 CD  cu O  O  Td  CD  Td cd  &  CJ Td  60  cj  S3 JH  cd  £  H  .2  +1  b  £3 O  00  b  T t  +1  CN  ON  "+l  CN T t  b  +1  ro;  O  +i  CO T t  b  +1  C N  00 •;  M  o 43  £3  i  i £  00  *o OH  Td  4 -—> CO  +1  e CD  I  OO  CN  i'r]  iri  r-~=  b  +1  u o CD  CD  NO  60  cd  M  CD : cd CD  43  ^4—»  . cd*  S  O CN  o  0  O CN  CN  CD  Iw3"  •  •  lill  *M  £ M CO  >.  1  O  UO CO  IT)  CD CD  CD Td  CM  CD 00  CO  £3  ">  i3 M  i  cd Td  CO  b  •B  l-l  cd  _  •©-  "3  Iii  CO T t  ^ij j j  CD  u  'I  CO  Td CJ >  CO  CO  b . .00 ' J *  co  <M  u  ;  00  m  T t  "o  £3  42  NO  £3  CJ  00  1  +1  C/f  j>  B,  PH O  '• :  CO _  3 o  ro b  •'.  VH  <M  SI  O V  B  CJ  O  •  ill  T t  CO  o o  < <  r-  b  +1  T t  42 IH  J  T-H  PH  NO  ro  CN T t  +1  T t  ta U  ON  +1  CN  o  00  43  op '53  o  •j  1^1  T t  NO  no  liO  10  CD  in  cd  160000  0  _i 0  ,  ,  ,  ,  10  20  30  40  1 50  Time (days) Figure 35  Degradation profiles of control PLG50:50 microspheres  with differing inherent viscosities (l-20pm). Molecular weights were measured by G P C at 4 0 ° C using polystyrenes as standards (Solvent: chloroform. Solvent flow rate: 1.0 mL/min).  95  120000  100000  I  80000 •PLG65:35 (I.V. =0.55)  £  •PLG75:25 (I.V. = 0.55)  60000  •PLG85:15 (I.V. = 0.56) •PLG100:0 (I.V. =0.60)  40000  20000 -I  1  1  1  1  1  0  10  20  30  40  50  T i m e (days)  Figure 36 20pm)  with  Degradation profiles o f control P L G microspheres (1different  LA:GA  compositions  and similar  inherent  viscosities. Molecular weights were measured by G P C at 4 0 ° C using polystyrenes as standards (Solvent: chloroform. Solvent flow rate: 1.0 mL/min).  96  100000  20000 -I  1  1  .  1  1  0  10  20  30  40  50 Time (days)  Figure 37 PLG  Degradation profiles of control and 10% paclitaxel-loaded  microspheres  (l-20u.ni)  with  different  LA:GA  compositions.  Molecular weights were measured by G P C at 4 0 ° C using polystyrenes as standards (Solvent: chloroform. Solvent flow rate: 1.0 mL/min).  97  100000  20000 -I 0  •  1  1  1  1  1  1  5  10  15  20  25  30  Time (days) Figure 38 PLG  Degradation profiles of control and 10% paclitaxel-loaded  microspheres  (l-20um)  with  different  LA:GA  compositions.  Molecular weights were measured by G P C at 4 0 ° C using polystyrenes as standards (Solvent: chloroform. Solvent flow rate: 1.0 mL/min).  98  PLG50:50 (I.V. = 0.78), 1-20um = 0.78), 3O-100um  I  40000 30000 -I 20000 0  10  15  20  25  30  Time (days)  Figure 39  Degradation profiles of control PLG50:50 (I.V.= 0.78,  1-  20pm & 20-100pm) microspheres. Molecular weights were measured by GPC  at 4 0 ° C using polystyrenes as standards (Solvent:  chloroform.  Solvent flow rate: 1.0 mL/min)  99  4.6.2. Surface morphology by SEM Figure 4 0 A - H shows the S E M morphologies of control PLG100:0 microspheres after incubation in PBS-albumin at 3 7 ° C over 6 weeks. Surface erosion was evident at day 3 and became more pronounced at day 28. A t day 35, the outer layer of the microspheres started to peel off on some of the microspheres. Figure 4 1 A - H shows surface morphologies of control PLG65:35 microspheres after incubation in PBS-albumin at 3 7 ° C . A t day 28, pinhole structures were found in some of the microspheres and there was evidence of bulk degradation in some of the microspheres at day 42. The surface morphologies of control PLG85:15 and PLG75:25 microspheres with similar inherent viscosities were similar to those of control PLG65:35 microspheres. Figure 4 2 A - H shows micrographs of control PLG50:50 microspheres with the lowest inherent viscosity of 0.58  degrading in PBS-albumin at 3 7 ° C over 6 weeks.  Microspheres showed relatively smooth surface morphologies until day 21, at which time pore structures appeared on some of the microspheres. The number of pores increased and at day 42, due to the extensive bulk degradation inside the microsphere matrix, the microspheres collapsed and an extensive pore network was observed. Figures 4 3 A - F are micrographs of 10% paclitaxel-loaded PLG50:50 (I.V.= 0.74, 1-20 pm) microspheres degraded in PBS-albumin at 3 7 ° C over 4 weeks. There was evidence of significant degradation of the microsphere matrix by day 28.  100  A . (Day 1)  B. (Day 2)  G . (Day 35)  H. (Day 42)  Figure 40  S E M micrographs of control PLG100:0 ( I . V . = 0.60,  20-  100pm) microspheres degrading in PBS-albumin at 3 7 ° C .  101  A.  (Day 1)  B. (Day 3)  C . (Day3)  D. (Day 14)  E . (Day 21)  F. (Day 28)  G . (Day 35)  H. (Day 42)  Figure 41  S E M micrographs o f PLG65:35  (I.V.=  0.55,  l-20um)  control microspheres degrading in PBS-albumin at 37°C.  102  A . (Day 1)  B. (Day 2)  G . (Day 35)  H. (Day 42)  Figure 42  S E M micrographs of PLG50:50  (I.V.= 0.58,  l-20um)  control microspheres degrading in PBS-albumin at 3 7 ° C .  103  A . (Day 1)  E. (Day 21)  Figure 43  B . (Day 3)  '  F. (Day 28)  S E M micrographs of PLG50:50 (I.V.= 0.74, l-20um) 10%  paclitaxel-loaded microspheres degrading in PBS-albumin at 3 7 ° C .  104  4.6.3. Glass transition temperatures of the degraded PLG microspheres Figure 44 shows a change in peak position o f the glass transition temperatures o f the control PLG50:50 (I.V.= 1.06) microspheres during degradation over 4 weeks in PBS-albumin buffer. Table 11 summarizes the change in glass transition temperatures o f all control microspheres during 28 or 42 days o f degradation in PBS-albumin. There was a consistent decrease in the glass transition temperatures for all P L G microspheres with a L A : G A composition o f 50:50. In contrast, the P L G polymer microspheres with different L A : G A compositions did not show any significant changes in their glass transition temperatures.  o  X  W D C B  o  53  A  35.00  40.00  45.00  50.00  55.00  60.00  65.00  Temperature (°C)  Figure 44 (I.V. 1.06,  Glass transition temperatures o f control P L G microspheres l-20um) with a L A : G A  ratio o f 50:50 after 28 days o f  degradation in PBS-albumin at 3 7 ° C (A: day 1, B : day 3, C : day 7, D : 14 and E : day 28).  105  ON  CN  b  Q  U  .S3  CL,  IT)  <U-'  «HJ—I .S  a  o  o .2  r-H  13  Q  ©  +1  oo .  ,Tt  (O  00  +1  ; 0 \ -;.  7\  Tt  Ti-  ro" in  CN  ro  ro  ON  +1  +1  "3. T3 CU  b  T—H  CN  .+1 ON  Q  '  oo TT  CN  NO  ON Tt  — m  1—1  ON  +1  in iCN | m  -d  8  •O • 00  <+H  s  > l-i  s  3 c2 "cd S  Q  Q  cd  3  cd  U  <u i n CO  b in  •  cd m £ H vo cd  +1  +1  ro ro' in  «•*-.  -r in  • OO b  00  oo. b •, +l  b  +1  +1  Tt  NO in in  oo  c-b  'j—i  r~  in m  -  in in in  b  *•  in p b  ! ro ! b  >r, CN  ;  ©  ;  — —<  cd  >  +i © m  '  CD  1 3  1 3  CN  s  ©  +1,; T t  CN in  :  i © ' (in 1 * ,  K M  ON  ©  +1  ON  '  '©'  ©  +1  CNJ  m' m  +1  in © •  CN 4H  +1  ON  O  ro in  m  J  co  +1 ocT  S3  o '"S S3  <D ^—> CD 13  CD CD  •a  o  cd  a H-l  a  o  in CN in  in oo  in  CD  CO  cS  in  NO  00  HI,..  CD N  <D  4-*  S3  CD  <u  4—>  -4—>  -*->  -H»  CO  CO  CO  VH  J—1  -M  <D 43  -4—»  IH 4—»  CD 43 -4—»  •4—»  cd -*cd -> -*cd ->  CD w  CD 42  43  60 O ' - o in iO b b in  CO  CD  m ©  CD  CD  »—H  and day  +l m  cd  00  Tt  ©•'"•  cd  3  O  in in  ro  3  CO  .a 8  .  CN  'cN . >n  ©  • 00 TT  +1 - +1  Q  b +l  ca  o  in  ';  CO  £co s c  CN  'TT  © '  co  s^ cd  ,  ,+l  1  5  +1  r,©'..  o o —  b  m  +-»  1  'b '+i CN ,—<  CN  si  ON Tf  in ' m ON  CO  I3 ! I.  • CD  •b i n •..  CN ro  I  nperaltures betAveen day  <+-*,.  ©  ion  H-l PM  13  Q  +1  +1  .  glass  a o  ON  iled -tests on  CD  oo b  a  to,.  I.CN  o o  o  CN  o  CN  JL  C/3 ;NO  ro  00  © ©  © CN  00  m b  liilSl jijjjll Bill! Illi  © CN  © CN  ©  CN  ;_1 1  ©  CN  i  1Jf  43  . • !.  © ©  Two-  S3  £  ©  and day  *  co  and day  H-l  'S ^  00  C  veen day  43  nperaltures bet\veen day  • <D  l-l  .2 -2 S "§)  nperaltures  b +l  oo OO CN CN CN T t  on  b +i  ro m  on  ON  .+1 '  m CN in  glass  b >/-.  ©  glass  >H  "  iled -tests on  +1  -H Tt  CD  00 <JH O c3  ©  b ' +l  cd  oo;-.  3  ro  oo b  ON'I  CN  Q  +1  o NO  NO Tt  0O  -d •  IH H00 J  +1  ro  60  'to  o  b  +1  cd  0  VO  iled •tests on  o to  ON Tt  m  Two-  I—1  Tt  <D •  43  in  t>  in co  £3  cd cd  oo  Two-  O  +1 m  Tt  o  co Id  +1  +1 NO  NO  in / © ••  m m  m •n  d  13  4.7.  The effect of y-irradiation on paclitaxel, control and paclitaxel-loaded microspheres The impact o f y-irradiation at a dose o f 2.5 Mrad on paclitaxel recovery and  paclitaxel  loading in the microspheres was determined.  There  are no significant  differences for either paclitaxel recovery or loading o f paclitaxel in microspheres before and after gamma irradiation. The effect o f gamma irradiation on the molecular weights o f PLG85:15 polymer in 20% paclitaxel-loaded microspheres showed that gamma  irradiation  caused a  significant decrease in molecular weight of the polymer. Table 12 summarizes the molecular weights o f non-irradiated control PLG85:15 microspheres and non-irradiated and gamma irradiated 20% paclitaxel-loaded PLG85:15 microspheres after incubation in PBS-albumin at 3 7 ° C . The degradation profiles are shown in Figure 45. Slopes of the linear regressions of these degradation curves revealed that the in vitro degradation rates o f control and 20%> paclitaxel-loaded microspheres (non-irradiated) were not significantly different, but that they were significantly different from the degradation rate of the gamma irradiated (2.5 Mrad) 20%> paclitaxel-loaded microspheres.  107  Table 12  Summaries o f the MGPC o f non-irradiated control PLG85:15  microspheres and non-irradiated and gamma irradiated (2.5 Mrad) 20% paclitaxel-loaded  PLG85:15  microspheres after  incubation  in P B S -  albumin at 37°C  Time (day)  MGPC of control and 20% paclitaxel-loaded PLG85:15 microspheres Control (xlOOO)  0  88.4+1.3  3  83.4 ± 1.1  83.4+1.3  14  85.9+ 1.3  17  88.2 + 1.2  24  92.0 ± 1.3  28  [  ' "  j  " 88.7 ± 1.4  E  88.2+1.2 :  91.7+1.4-  .'  50.7 ± 0 . 9 57.5 ± 0 . 9  90.0+1.3 1  88.2 ± 1 . 3  65.1 ± 0 . 9 63.0 ± 1 . 0  56.7 ±  •". -  0.9 ^ ^_ T  84.8 + 1.2  52.1 ± 0 . 8  85.8 + 1.3  '50.1 ± 0 . 8  87.9 ± 1.3  86.5 ± 1.3  52.1 ± 0 . 7  35"  83.4 ± 1.3  87.9 ± 1.1  45.2 ± 0 . 8  42  82.0 ± 1 . 3  80.2 ± 1 . 3  47.0 ± 0.7  83.611.3_:-  78.1 ± 1 . 3  44.9 ± 0.6  76.9 ± 1.4  87.1 ± 1.3  42.0 ± 0.7  83.4± i . i ;  40.4 + 0.6 31.9 ± 0 . 5  51  ".  58 ;  •  3  3  87.1 ± 1.4 10  After irradiation (x 1000)  Before irradiation (xlOOO).  8  74.0+1.2  |  |  -  72  78.1 ± 1.2  73.2 ± 1 . 3  79  76.fi'1.3  74.0 ±'V2T I""-;  a  v  " ;  30.9  ±  0  77T  .  5  Molecular weights (g/mole) o f control, paclitaxel-loaded PLG85:15 microspheres (before and after irradiation) are the averages o f three sample runs, ± standard deviation.  108  !  100000  20000 -i 0  1 20  1 40  1 60  1 80  1 100  T i m e (days) Figure 45 The degradation profiles in PBS-albumin at 3 7 ° C o f nonirradiated control and 20% paclitaxel-loaded PLG85:15 (I.V.= 0.56, 120pm) microspheres as well as 20%> paclitaxel-loaded PLG85:15 (I.V.= 0.56, l-20pm) microspheres after irradiation at a dose o f 2.5 Mrad. The slopes o f linear  regression of the degradation  profiles  were -0.16  (g/mole/day) for the control microspheres, -0.18 (g/mole/day) and -0.38 (g/mole/day)  for the paclitaxel-loaded PLG85:15 microspheres (before  and after irradiation respectively).  109  5.  DISCUSSION  5.1.  Composition of the P L G polymers The area under each N M R signal in the spectrum was proportional to the numbers  of hydrogen atoms in that group. The average molar compositions o f the polymers could be determined by the ratio o f the corresponding H ' - N M R integrated peak area (Amecke et al., 1995; Deasy et al., 1989). The chemical shift (5), which was the difference in absorption frequency between a given proton and the proton from the internal standard ( T M S ) , was used to identify the, existence o f certain protons. A peak at chemical shift (8) = 7.2 ppm was due to the impurity from the solvent CDCI3. The lactide unit had two different protons: methine and methyl. These proton peaks appeared at 5.2 ppm and 1.5 ppm respectively (Asano et al., 1989). The methyl proton appeared as a doublet peak and the methine proton was a quartet peak, due to spin coupling interactions  between  neighboring protons, and was related to the number of possible spin orientations that these neighbors could possess, or so-called spin-spin splitting or spin coupling. The methylene proton in the glycolide unit (Figure 7) had a multiple peak at 4.8 ppm, which may be due to vicinal coupling (Schakenraad, et al,  1989; Amecke et al,  1995). The  additional small peaks at 1.3 ppm could be from residual catalyst in the polymer (Amecke et al., 1995). Our data confirmed the molar ratios o f L A : G A in the polymers received from BPI.  5.2.  Microsphere fabrication  5.2.1. Size and surface morphology of the microspheres The solvent evaporation method has been commonly used in manufacturing many controlled release formulations for hydrophobic drugs with high encapsulation efficiency  110  (Ruiz et al,  1991; Alonso et al,  1993; O'Hagan et al,  1994 and Jeffery et al,  1991).  Acceptable amounts of residual solvent ranging from 8.8 ± 1.2 to 463.1 ± 20.8ppm have been reported using this method (Bitz et al,  1996).  Factors affecting the size ranges of microspheres manufactured have been widely investigated. These factors include polymer concentration, drug concentration, ratio of organic/aqueous phase volume, concentration of P V A solution and rate of agitation (Jeffery et al,  1991; Conti et al, 1995 and Nihant et al,  1995). The concentration of the  P V A solution and the rate of agitation during emulsification have been shown to have a profound influence on the sizes of the microspheres (Jeffery et al,  1991). Increasing the  P V A concentration or increasing the agitation rate decreases the sizes o f the microspheres by causing an increased shear stress on the organic phase droplets and decreasing the size of the dispersed phase. Two size ranges of microspheres were produced in these studies. Microspheres with sizes of less than 20pm were prepared using 5% (w/v) P V A solution and a stirring rate of 900 ± 5 0 rpm at room temperature. Microspheres with size ranges between 20-100 p m were prepared by changing the P V A concentration and stirring rate to 2.5% (w/v) and 550 ± 50 rpm, respectively. Microspheres produced were spherical, with sizes controlled within the targeted ranges. The results indicated that polymers with either different molecular weights but the same L A : G A  composition or different L A : G A  compositions did not produce  microspheres with any significant differences in their size distributions. A l l small microsphere batches were less than 20 p m , whereas the large microsphere batches were within 20-100 pm.  Ill  Scanning electron micrographs showed that all 3 batches of the 20% paclitaxelloaded PLG100:0 (I.V.= 0.60,  20-100pm) microspheres (Figure  14A) possessed an  extensive pinhole-like network on the microsphere surfaces. However, this effect was not observed in the scanning electron micrographs of the 10% paclitaxel-loaded (Figure 15 A ) and control PLG100:0 (Figure 16A) microspheres. Bodmeier et al. (1987) reported that the surfaces of quinidine-loaded PLG100:0 microspheres changed from virtually smooth at a low drug loading of 7% (w/w)  to a pinhole-like structures at a loading of 24% (w/w).  The pinholes became more evident as the loading of quinidine was increased from 34 to 62%o. A porous substructure in the microsphere matrix was found under this pinhole-like skin structure on the microspheres (Bodmeier et al,  1987). Similar results were also  reported i n a hormone-loaded P L G 7 5 : 2 5 microsphere system b y T a k a d a et  al.  (1994). These pinhole structures were not seen on other 10 and 20% paclitaxel-loaded microspheres in which glycolide was present in the polymer chain (Figure 14, 15 and 16). Several factors, which include the rate of solvent removal, the compositions of the polymer  wall material and the  loading of paclitaxel,  may  influence  the  surface  morphology of the microspheres. A lower solubility of paclitaxel in the PLG100:0 polymer microspheres may have led to precipitation of the drug in the superficial layers of the microsphere surface. The pinhole network might then form following the washing and drying steps of microsphere manufacture.  5.2.2. X-ray diffraction patterns of control and paclitaxel loaded microspheres X-ray diffraction data for all microsphere samples showed only an amorphous matrix. There was no evidence of a crystalline form of paclitaxel present in the  112  microspheres. Paclitaxel may have been present in the polymer matrix either as a solid solution and/or in an amorphous form. Scanning electron micrographs of sectioned microspheres did not provide any evidence of crystalline paclitaxel present in the bulk microsphere matrix. In addition, D S C analysis of microsphere samples showed an absence o f any melting peaks due to crystalline paclitaxel.  5.2.3. Paclitaxel total content and encapsulation efficiency The encapsulation efficiencies of paclitaxel in P L G microspheres were found to be independent of the paclitaxel loading or polymers. Except for the 20% paclitaxelloaded P L G 100:0 microspheres, all P L G microspheres showed over 90%> encapsulation efficiency. This is comparable with many other P L G microsphere systems, which have reported encapsulation efficiencies of about 90-95%> for hydrophobic drugs (Okada et  al,  1995). The reason for PLG100:0 20%> paclitaxel-loaded microspheres having the lowest encapsulation efficiency (86.5%>) among the microspheres could be related to the pinhole structure seen in the scanning electron micrographs (Figure 14A). Paclitaxel may have been lost from the outer superficial layers either during the fabrication process and/or during the final washing step.  5.2.4.  Glass transition temperatures of control and paclitaxel-loaded microspheres D S C data for control and paclitaxel-loaded microspheres revealed that the glass  transition temperatures increased with an increase in the paclitaxel loading, regardless of the sizes of the microspheres, molecular weights and compositions of the polymers. However, there was no increase in the T  g  of the control P L G microspheres/paclitaxel  physical mixture (Figure 23B), suggesting that the increase in the glass transition temperatures were related to the presence of paclitaxel in the microsphere matrix. Similar  113  increases in glass transition temperatures were also reported by Yamakawa et al. (1992) for neurotensin-loaded P L G microspheres. The increase in T  g  could be due to either a  reduction o f the free volume in the polymer matrix after the incorporation o f paclitaxel, or an interaction between the paclitaxel and the polymers. This interaction would likely make the polymer chains more rigid and require more energy to achieve segmental motion. Okada et al. (1995) speculated that drug molecules in the microspheres would be dispersed throughout the P L G matrix and that the motion o f P L G segments would be hindered by the interaction between the polymer and drug. Maulding (1987) and Prinos et al. (1995) have suggested that drug-polymer interactions might be due to intermolecular hydrogen bonds in the polymer-drug matrix. Menikh et al. (1993) and L i n et al. (1993 & 1995) used Fourier transform infrared spectroscopy (FTIR) to characterize drug-polymer solid-state interactions and confirmed the existence o f intermolecular hydrogen bonds.  5.2.5. Investigation of polymer-paclitaxel interaction by Fourier transform infra-red spectroscopy (FTIR) The C = 0 stretching vibration is the most characteristic band o f the C = 0 group, which absorbs very strongly in the region 1750 ± 80 cm" . However, the peak position o f 1  the absorbance band is greatly affected by the adjacent group and the surrounding environment such as C O 2 , making it unsuitable for detecting any subtle changes due to hydrogen bonds (Ishida, 1987). The deformation o f the C = 0 bond as part of the ester group can be detected in the region 710 ± 80 cm" , which has been observed for many organic compounds and has 1  been attributed to in-plane and out-of-plane deformation o f the C = 0 bond (Roeges, 1995). The 5 C = 0 (in-plane deformation) and y C = 0 (out-of-plane deformation) would  114  appear as one peak i f there was no influence from outside molecules (Roeges, 1995). Figure 25 A showed that only one absorbance band existed at 710 cm" for both the 8 C = 0 1  (in-plane  deformation)  absorbance  bands  and y C = 0 were  (out-of-plane  observed  for  deformation)  control  for  microspheres  paclitaxel. and  Two  control  microspheres/paclitaxel physical mixture as well as paclitaxel-loaded microspheres. The peak with lower wavenumber was identified as C = 0 deformation since it was intensified in the F T I R spectra of paclitaxel-loaded microspheres and physical mixture of control microspheres/paclitaxel. Therefore, the C = 0 deformation peak for control microspheres was 708 cm" (Figure 25B). The peak with higher wavenumber was relatively stable 1  regardless of polymer compositions. A significant shift of the C = 0  deformation band was observed for the  20%  paclitaxel-loaded microsphere sample (722 cm" ) (Figure 26B). This change in the band 1  position implies that the required energy for C = 0 deformation was increased due to the presence of paclitaxel in the polymer matrix. It is possible that hydrogen bonds between the hydroxyl group of paclitaxel and the ester group of the polymer may be the major cause of the shift of the C = 0 deformation band. Intermolecular hydrogen bonds between paclitaxel and P L G polymer chains could result in the loss of some degree of flexibility, which would increase the energy required for the glass transition. This may explain the increase in glass transition temperatures with an increase in paclitaxel loading in the microspheres.  115  5.2.6. In vitro degradation of microspheres 5.2.6.1. Molecular weights and morphological changes The P L G polymers with different lactide:glycolide compositions possessed very similar inherent viscosities. Except for PLG65:35 which had a molecular weight of about 46,000 g/mole, this group of polymers had molecular weights between 82,000 and 96,000 g/mole. After 28 days of incubation in the PBS-albumin at 3 7 ° C , slight decreases (between 3-13%) in the molecular weights were observed in this group of polymers, either with or without paclitaxel loading. Ramchandani, et al. (1997) also reported a slow degradation of a PLG85:15 implant with a M  w  of 90,801 g/mole in phosphate buffer (pH  = 7.4) in the first 63 days. A period of slow degradation in phosphate buffer (pH = 7.4) has been reported for most P L G microspheres with higher lactide content (> 50 mole%) by Park (1995). Microspheres loaded with or without paclitaxel had almost identical degradation profiles (Figures 37, 38 and 45), indicating that the presence of paclitaxel in the polymer matrix had little effect during the first four to six weeks on the degradation of this group of P L G polymers. S E M micrographs showed very little or no changes in the surface morphologies of PLG85:15, 75:25 and 65:35 microspheres over 4-6  weeks,  providing additional evidence of slow degradation over this time period. The period of slow degradation is always accompanied by little mass loss as reported by Fukuzaki et al. (1991). Amorphous poly(/-lactide-co-glycolide) polymers with 70 mole% of/-lactide and molecular weights of 16,900, 24,000 and 41,300 g/mole showed little change in the masses following implantation in rats for about 5 weeks. Poly(Z-lactide-co-glycolide) containing 85 mole% o f /-lactide ( M  w  = 114,000 g/mole)  showed minimal mass loss in the first 10 weeks in phosphate buffer (Vert et al., 1991).  116  The in vitro degradation profiles of P L G polymers with 80 mole% ( M = 54,000 g/mole) w  and 90 mole% ( M  w  = 19,200 g/mole)  o f d,/-lactide  content were alike, but were  significantly different from a P L G polymer containing 50 mole% ( M = 48,500 g/mole) w  of d, /-lactide content (Park, 1995). The group o f P L G polymers with the same composition o f glycolide and lactide (50:50) possessed very different inherent viscosities. The molecular weights were about 140,000 g/mole for PLG50:50 (I.V.= 1.06), 71,000 g/mole for PLG50:50 (I.V.= 0.78), 63,000 g/mole for PLG50:50 (I.V.= 0.74), and 46,000 g/mole for PLG50:50 (I.V.= 0.58). During 28 or 42 days o f degradation in the PBS-albumin at 3 7 ° C , the molecular weights of this group o f P L G polymers decreased significantly, regardless o f paclitaxel loading. Degradation of the polymers was generally slow in the first three days and then occurred more rapidly. B y day 28, the molecular weights had decreased to about 18 to 43% of their initial molecular weights. Polymers with higher initial molecular weights had a greater reduction than the polymers with lower initial molecular weights in the same period o f time. This was because polymers with longer chains possess a greater number o f sites for hydrolysis, leading to a faster decrease in the molecular weight. Similar results were reported by Shah et al. (1992), who also employed 50:50 P L G polymers in the studies. There appeared to be no evidence o f changes in the surface morphology o f PLG50:50 microspheres until about 21 days o f incubation in PBS-albumin (Figure 42). B y day 28, S E M micrographs showed evidence o f significant mass loss. Reed and Gilding (1981) also showed that mass loss did not occur until after 21 days o f in vitro degradation of PLG50:50 polymers.  117  The higher degradation rates for PLG50:50 polymers are due to the higher glycolide content which is more hydrophilic than the lactide component (Park, 1995). The degradation half-life of a PLG50:50 implant ( M  w  « 46,000 g/mole) in Sprague-  Dawley rats was about one week, while the degradation half-life of a PLG100:0 implant was  about  6 months  microspheres ( M  w  «  (Miller  et  al,  1977). Complete  degradation  of  PLG50:50  15,500 g/mole) was seen after 63 days implantation in male  Sprague-Dawley rats (Visscher et al,  1985).  Figure 29 compared the degradation profiles of two different sizes of PLG50:50 (I.V.  = 0.78)  control microspheres. The rate o f degradation was slow for the larger  microspheres for the first 10 days, followed by more rapid degradation. The smaller microspheres showed a steady drop in molecular weight similar to the PLG50:50 (I.V.  =  0.74) microsphere degradation profile (Figure 28). The slopes of the two degradation curves indicated that there was no significant difference between the overall degradation rates. Visscher et al. (1988) also reported that the degradation profiles o f PLG50:50 microspheres ( M W « 43,000 g/mole) with three size ranges (45-75 p m , 75-106 p m and 106-177 pm) were similar. The degradation of P L G microspheres involves the uptake o f water into the polymer matrix, random chain scission of the linkage of ester bonds in the polymer backbone, decrease in molecular weight of the polymer and ultimately mass loss from the microspheres. It has also been suggested that degradation of P L G microspheres occurs more rapidly in the centre than at the surface due to the autocatalytic action of the carboxylic acid end groups o f degrading P L G polymers (Li et al,  1990A and 1990B;  Therin et al., 1992). Similar results were also seen in our studies (Figure 42).  118  5.2.6.2. The effect of degradation on glass transition temperatures There was a small but consistent drop in the glass transition temperatures for all P L G microspheres with the same 50:50 L A : G A composition after incubation in P B S albumin. P L G microspheres with different L A : G A compositions, in contrast, did not show any significant changes in their glass transition temperatures. Shah et al. (1992) reported a similar decrease in glass transition temperature associated with a decrease o f molecular weight o f a PLG50:50 polymer. According to the Flory-Fox equation, a decrease in the molecular weight o f a polymer would lead to a decrease in its glass transition temperature. The drop in glass transition temperatures o f P L G polymers with the same 50:50 L A : G A composition was likely related to the change in molecular weights. For the P L G polymers with differing L A : G A  compositions, no significant  changes in glass transition temperatures were observed up to 28 or 42 days of degradation in PBS-albumin at 3 7 ° C since the changes in the molecular weights in this group o f microspheres were small.  5.2.7. In vitro release ofpaclitaxel from paclitaxel-loaded microspheres The solubility of paclitaxel in the PBS-albumin was about 3 p g / m L (Winternitz, 1997  and Zhang 1997). Sampling intervals were selected in order to prevent the  concentration o f paclitaxel in the release buffer from exceeding 15% o f its aqueous solubility. This was to ensure that sink conditions were maintained during the entire course o f the in vitro release studies (Carstensen, 1977). Nevertheless, due to the rapid release o f paclitaxel  during the burst phase, sink conditions may not have been  maintained in the first 3 days, particularly for the 20% paclitaxel-loaded microspheres.  119  However, this would not be expected to alter the cumulative amount of paclitaxel released in the course of the study. The release profiles of paclitaxel from PLG50:50 microspheres with paclitaxel loading from 1 to 20% (Figure 27) showed that a significant amount of paclitaxel, which accounted for about 10-20%o of total paclitaxel loading, was released during the first two or five days. The initial burst phase is believed to be due to the large amount of drug released from the superficial surface layers of the microspheres (Alonso et al,  1993).  Paclitaxel was released at a relatively steady rate between day 5 and day 30 until the secondary burst phase appeared at day 31 for the 20% paclitaxel-loaded microspheres (Figure 27). The microspheres with higher paclitaxel loading provided faster paclitaxel release due to the higher concentration of paclitaxel in the microsphere matrix. According to the kinetic theory for diffusion controlled release from a sphere, the amount of drug released at a given time is directly proportional to the drug loading (Baker, 1987). Paclitaxel release  from  the microsphere matrices was probably due to  the  combination of diffusion and degradation process. Diffusional control is usually the major mechanism of drug release in the early phase of release (Heya et al.,  1991).  Degradation controlled release of drug would play a more important role at the later phase of drug release (Shah et al,  1992). There is evidence of a triphasic pattern of drug  release for the PLG50:50, 20%) paclitaxel-loaded microspheres (Figure 27) due to the appearance of a secondary burst phase at about day 31. This is likely due to the contribution of significant degradation to the drug release rates. Sturesson et al. (1993) reported a similar triphasic release pattern in PLG50:50 (I.V.= 0.2) microspheres loaded with timolol maleate. Alonso ei al. (1993) and Shah et al. (1993) also showed this release  120  pattern by using different model compounds (tetanus vaccine and red dye) encapsulated in PLG50.50 microspheres ( M  = 100,000 g/mole and I.V.= 0.4-0.5).  w  The 20% paclitaxel-loaded PLG100:0 microspheres showed a much higher initial burst phase of release compared to 10%> paclitaxel-loaded PLG100:0 microspheres (Figure 28). The extensive pinhole-like structures on the microsphere surfaces might be related to the high burst phase from the 20%> paclitaxel-loaded PLG100:0 microspheres. Bodmeier and M c G i n i t y (1987) showed that the burst phase of quinidine release from PLG100:0 microspheres increased from zero to almost 100%. when the loading of quinidine in the microspheres increased from 7 to 53% (w/w). Extensive pores were found on the microsphere surfaces when the loading of quinidine was over 24%> (w/w). Similar results were also reported by Takada et al. (1994) on thyrotropin-releasinghormone (TRH) loaded PLG75:25 microspheres ( M  w  = 14,000). Extensive pores formed  at high T R H loading, contributed to the high initial burst phase. The cause o f the apparent phase of faster release at about day 6 for the 20%>  paclitaxel-loaded  PLG100:0  i  microspheres is not clear. It may be related to the pinhole surface morphology of these microspheres. Fong et al., (1986) reported a similar pattern of faster release after the burst phase on a 44% thioridazine-loaded PLG100:0 microsphere system. The amounts of paclitaxel released in the first 15 days from PLG85:15, 75:25 and 65:35 microspheres were almost identical (Figures 29 and 30). Different lactide:glycolide compositions did not result in any differences in the paclitaxel release rates in the early phase of release. However, the release rates of PLG85:15 microspheres with 10%> and 20% paclitaxel loading started to decrease after day 15 compared with the PLG75:25 and PLG65:35 microspheres. Spenlehauer et al. (1988) also showed that the presence of  121  different  amounts o f glycolic units in the polymer backbones for PLG75:25 and  PLG90:10 microspheres loaded with 30% cisplatin had no influence on the drug release. No  differences in the paclitaxel release rates  from  10% paclitaxel-loaded  PLG50:50 (I.V.= 0.78) and PLG50:50 (I.V.= 0.74) microspheres were found in the first 24 days of release (Figure 31). However, the paclitaxel release rates from 20% paclitaxelloaded  PLG50:50  (I.V.=  significantly different  0.78) and PLG50:50  (I.V.=  0.74) microspheres  after day 10. The paclitaxel release rates  from  were  10 and 20%  paclitaxel-loaded PLG50:50 (I.V.= 1.06) microspheres were significantly lower than the release rates from the PLG50:50 microspheres with lower inherent viscosities o f 0.74 and 0.78. The slow release rate for paclitaxel from the higher molecular weight PLG50:50 (I.V.= 1.06) microspheres was likely due to decreased diffusion o f paclitaxel through the polymer and increased hydrophobicity of the matrix due to fewer carboxyl end groups in the higher molecular weight polymers. Fong et al. (1986) reported that thioridazine released from P L G 100:0 microspheres with different molecular weights but similar loading and size ranges were almost identical in the first 6 days. The release rate o f thioridazine from microspheres with the lower molecular weight increased after day 6. Heya et al. (1991) also showed that higher molecular weight P L G microspheres gave lower release rate o f T R H . The release profiles of two different sizes (1-20 p m and 20-100 pm) o f 10% paclitaxel-loaded PLG50:50 (I.V.= 0.78) microspheres (Figure 33) showed that the larger size microspheres had a burst phase which released about 5% o f paclitaxel from the total loading, while the burst phase o f smaller microspheres was equivalent to about 10-12% of paclitaxel total loading. The slightly higher cumulative amounts o f paclitaxel released  122  from the smaller microspheres may have been due to a higher burst phase of release, which was related to the increased surface area of the smaller microspheres. The slopes df the two release-curves after the burst phases were not significantly different. Visscher et al. (1988) also found that drug release rates from 9% ergot-alkaloid loaded PLG50:50 microspheres (45-75 p m , 75-106 p m and 106-177 pm) were similar. C h a and Pitt (1988) also found that drug (L-methadone) release rate was independent o f the sizes of P L G microspheres.  5.2.8. Effects of gamma irradiation on paclitaxel and PLG polymers No gamma  paclitaxel  irradiation.  degradation The  occurred in paclitaxel-loaded  molecular  weight  of  20%  microspheres after  paclitaxel-loaded  PLG85:15  microspheres dropped significantly after gamma irradiation. The rate o f degradation of the gamma irradiated 20%> paclitaxel-loaded PLG85:15 microspheres in PBS-albumin was increased compared with non-irradiated control PLG85:15 microspheres and 20%> paclitaxel-loaded microspheres. Volland et al. (1994) reported that gamma irradiation of polyester could result  in  simultaneous chain scission and cross-linking. Gamma  irradiation caused the formation of free radicals in the P L G polymer chains, which either underwent a chain scission to form smaller molecular weight polymer chains, or two radicals combined to form a cross-linked P L G polymer. However, an irradiation dosedependent decrease in molecular weight was usually reported (Volland et al.,  1994;  Hausberger et al., 1995). Spenlehauer et al. (1988) showed that gamma irradiation had minimal influence on drug release profiles and the release of cisplatin from PLG75:25 microspheres was found to be independent of the doses of gamma irradiation.  123  5.3.  Intra-articular microsphere formulation and future work The overall objective  of the project  was to  develop a paclitaxel  loaded  microsphere formulation suitable for prolonged intra-articular drug delivery. The ideal properties of the formulation include biocompatibility with joint tissue, injectability and controlled release of drug over a period of about 3 months. We believe that of the P L G polymers studied in this work, the P L G polymer with 50:50 mole % L A : G A provides the most suitable matrix for the formulation since the degradation lifetime of the PLG50:50 polymer is the most appropriate. Drug release and degradation of the polymer matrix would likely be complete in approximately 3 months. Future studies would involve the determination of the tolerability and efficacy of control and paclitaxel loaded microsphere formulations in animal models. For example, the total amount or dose of microspheres that could be safely administered into joints is not known. Therefore, a series of in vivo tolerability studies in animal models would be required. Increasing doses of control microspheres would be injected into animal joints to determine an appropriate dose range for the microspheres, which could be tolerated in animal joints. Once the maximum dose of microspheres is established, efficacy studies with paclitaxel loaded PLG50:50 microspheres would be undertaken in animal models of rheumatoid arthritis.  124  6.  SUMMARY AND CONCLUSIONS  6.1.  Control and paclitaxel-loaded microspheres with two different size ranges of 120pm and 20-100pm were prepared using the solvent evaporation method. The microspheres thus obtained had smooth and spherical shapes except for the 20% paclitaxel-loaded P L G 100:0 microspheres which showed a pinhole network.  6.2.  D S C results  indicated' that there was  an increase in  the  glass  transition  temperatures for the paclitaxel-loaded microspheres compared to control P L G microspheres. 6.3.  FTIR  results suggested that the glass transition temperature  increase in the  paclitaxel-loaded microspheres might be due to the formation of hydrogen bonds between paclitaxel and P L G polymers. 6.4.  Degradation studies of control and paclitaxel-loaded microspheres indicated that the molecular weights o f P L G with 50:50 lactide:glycolide deceased rapidly following incubation in PBS-albumin, while the molecular weights of higher lactide  content  (>  50  mole%> of lactide)  P L G polymers did not decrease  significantly until after 3 weeks of incubation. This was due to the  greater  hydrophilicity of the PLG50:50 polymers. 6.5.  In  general,  the  release  profiles  of paclitaxel  from  all  PLG  microsphere  formulations showed a burst phase of release, followed by a phase o f slower release. The burst phase was caused by rapid release o f paclitaxel from the superficial surface layers of the microspheres.  125  6.6.  The release rates of paclitaxel from PLG50:50 microspheres were influenced by paclitaxel loading and molecular weights of the PLG50:50 polymers. Increased loading and decreased molecular weight led to faster paclitaxel release rates.  6.7.  P L G microspheres prepared from polymers with L A : G A ratios of 85:15, 75:25 and 65:35 showed minimal effect on paclitaxel release rates.  6.8.  The two size ranges of microspheres showed minimal effects on the rates of paclitaxel releases from the microspheres.  6.9.  Gamma irradiation had no influence on the paclitaxel encapsulated in P L G microspheres. The molecular weight of P L G polymers decreased significantly after gamma irradiation. The degradation rate of the P L G microspheres was also affected significantly after gamma irradiation.  126  REFERENCES Taxol: a history of pharmaceutical development and current pharmaceutical concerns, Journal of the National Cancer Institute Monographs, 15, 141-147 (1993).  Adams J D , Flora K F \ Goldspiel B R , Wilson J W , Arbuck S G and Finley R:  Determinants of release rate of tetanus vaccine from polyester microspheres, Pharmaceutical Research, 10(7), 945-953 (1993).  Alonso M J , Cohen S, Park T G , Gupta R K , Siber G R and Langer R:  Synthetic resorbable polymers based on glycolide, lactide, and similar monomers. In Wise D L et al. (Eds): Encyclopedic Handbook of Biomaterials and Bioengineering, Part A, 2, 9771007 (1995).  Amecke B , Bendix D and Entenmann G :  Taxol (paclitaxel) in breast cancer, Hematology - Oncology Clinics of North America, 8, 121-40 (1994).  Arbuck S G , Dorr A  and Friedman M A :  Asano M , Fukuzaki H , Yoshida M , Kumakura M , Mashimo T , Yuasa H , Imai K , Yamanaka H and Suzuki K: In vivo characteristics  of low molecular weight copoly(/-lactide acid/glycolic acid) formulations with controlled release of luteinizing hormone-release hormone agonist, Journal of Controlled Release,  £ 777-722 (1989).  Diffusion-controlled systems. In Baker R (Eds): Controlled release of biologically active agents, Wiley, New York, 39-83 (1987).  Baker R:  thermographic and clinical comparison of three intra-articular steroid preparations in rheumatoid arthritis, Ann. Rheum. Dis., 38(1), 36-39 (1979).  Bird H A , Ring E F and Bacon P A : A  Influence of the preparation method on residual solvents in biodegradable microspheres, International Journal of Pharmaceutics, 131 171181 (1996).  Bitz C and Doelker E :  Ill  and McGinity JW: The preparation and evaluation of drugcontaining poly(Z)L-lactide) microspheres formed by the solvent evaporation  Bodmeier R  method, Pharmaceutical  Research, 4(6), 465-471 (1987).  Regression of collagen-induced arthritis with taxol, a microtubule stabilizer, Arthritis & Rheumatism 37(6), 839-845  Brahn E , Tang C and Banquerigo M L :  (1994).  Burt H M , Jackson J K , Bains S K , Liggins R T , Oktaba A M C , Arsenault A L and  Controlled delivery of taxol from microspheres composed of a blend ethylene-vinyl acetate copolymer and poly(</,/-lactic acid), Cancer  Hunter W L :  Letters 88, 73-79 (1995).  Cytotoxicity of taxol in vitro brain tumors, Cancer Chemother.  Cahan M A , Walter K A , Calvin O M and Brem H :  against  human  Pharmacol,  Caplow  MC,  and  rat  malignant  33, 441-444 (1994).  Shanks J and Ruhlen R:  disassembly, The Journal of Biological  Carstensen JT:  How taxol modulates  microtubule  Chemistry, 269(38), 23399-23402 (1994).  Pharmaceutics of solids and solid dosage forms, Wiley Interscince,  New York, pp.64 (1997).  and Thies C : The formation and characterization of hydrocortisone-loaded poly(±-lactide) microspheres, Journal of Pharmacy and  Cavalier M , Benoit IP  Pharmacology,  38, 249-53 (1986).  Uptake of P32-labelled cyclophosphamide from arthritis knee joints, Ann. Rheum. Dis., 28(6). 624-629  Cawley MI, McAllister J M , Thould A K and Balme H W :  (1969).  one-week subdermal delivery system for L-methadone based on biodegradable microcapsules, Journal of'ControlledRelease, 7, 69-78  Cha Y and Pitt C G : A  (1988).  Cleland J L , M a c A , B o y d B , Yang J , Duenas E T , Yeung D, Brooks D, H s u C , C h u H , M u k k u V and Jones A J :  The stability of recombinant human hormone in  128  poly(lactic-co-glycolic acid) ( P L G A ) microspheres, Pharmaceutical  Research,  14(4), 420-425 (1997).  Investigation on process parameters involved in polylactide-co-glycolide microsphere preparation, Drug Development and  Conti B , Genta I and Pavanetto F:  Industrial Pharmacy, 21(5), 615-622 (1995).  Craig P H , Williams J A , Davis K W , Magoun A D , Levy A J , Bogdansky S and Jones  JP Jr.: A biologic comparision of polylactin 910 and polyglycolic acid synthetic absorbable sutures, Surg. Cynecol. Obstet., 141(1). 1-10 (1975).  Cutright D E , Beasley JD and Perez B:  Histologic comparision of polylactic and  polyglycolic acid sutures, Oral. Surg., 32(1). 165-173 (1971).  De Silva M , Hazleman BL,!Thomas D P P and Wright P:  Liposomes in arthritis: a  new approach, The Lancet, 23, 1320-1322 (1979).  and Meegan M J : Preparation and characterization of lactic/glycolic acid polymers and copolymers, Journal of Microencapsulation,  Deasy P B , Finan M P  6(3),  369-378(1989).  Denis J N , Correa A and Greene A E :  A n improved synthesis of the taxol side chain  and of RP56976, Journal of Organic Chemistry, 55, 1957-1959 (1990).  Dingle J T , Gordon J L , Hazleman B L , Knight C G , Thomas D P , Philips N C , Shaw, IH, Fildes F J T , Oliver J E , Jones G , Turner E H and Lowe J S :  Novel treatment for  joint inflammation, Nature, 271(5643). 372-373 (1978).  Dordunoo S K and Burt H M :  Solubility and stability of taxol: effects of buffers and  cyclodextrins, International Journal of Pharmaceutics,  133. 191-201 (1996).  Phase II study and long-term follow-up of patients treated with taxol for advanced ovarian  Einzig A , Wiemik P H , Sasloff J, Runowicz C D and Goldgerg G L :  adenocarcinoma, Journal of Clinical Oncology, 10(11), 1748-1753 (1992).  129  Emkey R D , Linsay R, Lyssy J, Weisberg JS, Dempster D W and Shen V : The  systemic effect of intra-articular administration of cortisteroid on makers of bone formation and bone resorption in patients with rheumatoid arthritis, Arthritis Rheum., 39(2), 277-282 (1996).  Finley R S and Rowinsky E K :  Patient care issues: the management of paclitaxel-  related toxicities, The annals of Pharmacotherapy,  Flory PJ:  28, S27-S34 (1994).  Morphology of the crystal state in polymers, Journal of Am. Chem. Soc,  84, 2857-2867 (1962).  Flory PJ:  Principles of polymer chemistry, Cornell University Press, 1953.  Evaluation of biodegradable microspheres prepared by solvent evaporation process using sodium oleate  Fong JW, Nazareno JP, Pearson J E and Maulding H V :  as emulsifier, Journal of Controlled Release, 3, 119-130 (1986).  Foong W C and Green K L :  liposome-entrapped  Treatment of antigen-induced arthritis in rabbits with methotrexate injected intra-articularly, Journal of  Pharmacy & Pharmacology,  Fox  45(3), 204-209 (1993).  T G and Flory PJ: Second-order transition temperatures and related properties of polystyrene: (I) Influence of molecular weight, Journal of Appl.  Physics, 2± 581-591 (1950).  Fukuzaki H , Yoshida M , Asano M , Kumakura M , Mashimo T , Yuasa H , Imai K and  In vivo characteristics if high molecular weight copoly(Xlactide/glycolide) with S-type degradation pattern for application in drug  Yamanaka H :  delivery systems, Biomaterials,  12, 433-437 (1991).  Crystal structure of 2-debenzoyl, 2-acetoxy paclitaxel (taxol®): Conformation of the paclitaxel side-chain, Pharmaceutical  Gao Q , W e i J M and Chen S H :  Research, 12(3), 337-341 (1995).  D K and Reed A M : Biodegradable polymers for use in surgery polyglycolic/poly(actic acid) homo- and copolymers: 1, Polymer, 20, 1459-  Gilding  1464 (1979).  130  Mechanisms of polymer degradation and erosion, Biomaterials, 17(2), 103-114 (1996).  Gopferich A :  Hydrogels and biodegradable polymers for controlled delivery of drugs, Polymer News, 8, 230-236 (1982).  Graham N B and W o o d D A :  Grem  J L , Tutsch K D ,  Simon K J Alberti  DB,  Willson  J K V , Tormey  DC,  Swaminathan S and Trump D: Phase I study of taxol administered as a short I V infusion daily for 5 days, Cancer Treatemnt Reports, 71(12). 1179-1184 (1987).  Pathophysiology and implications for therapy. In: Mechanisms of disease. In Epstein F H (Eds): Rheumatoid arthritis, New England Journal of Medicine, 32Z 1277-1289 (1990).  Harris E D :  Harris E D :  Rheumatoid arthritis, W.B. Saunders Company, (1997).  and Deluca PP: Gamma irradiation effects on molecular weight and in vitro degradation of poly(DL-lactide-co-glycolide) microparticles, Pharmaceutical Research, 12(6). 851-856 (1995).  Hausberger A G , Kenley R A  Biodegradable polymers in controlled drug delivery, Crit. Rev. Ther. Drug Carrier Syst, 1(1), 39-90 (1984).  Heller J:  Controlled drug release from poly(ortho esters) - a surface eroding polymer, Journal of Controlled Release, 2, 167-177 (1985).  Heller J:  Controlled release of biologically active compounds from bioerodible polymers, Biomaterials, ]_, 51-57 (1980).  Heller J:  X-ray studies of the structure of gelatin micelles, Z. Physik. Chem., 10, 371-394 (1930).  Herrmann K, Gemgross O and Abitz W :  131  Factors influencing the profiles of T R H release from copoly(DL-lactic/glycolic acid) microspheres, International Journal of Pharmaceutics, 72, 199-205 (1991).  Heya T , Okada H , Ogawa Y and Toguchi H:  Methotrexate therapy in rheumatoid arthritis: experience, American Journal of Medicine, 75 (6A). 69-73 (1983).  Hoffmeister  RT:  15-years  Biodegradable Polymers, Advances in Pharmaceutial Sciences, Academic Press, 101-164 (1992).  Holland SJ and Tighe B J :  Holland SJ, Tighe B J and Gould P L : Polymers for biodegradable medical devices.  1. The potential of polyesters as controlled macromolecular release systems, J. Controlled Release, £ 155-180 (1986).  Granulocyte survival in synovial exudate of patients with rheumatoid arthritis and other inflammatory joint disease, Yale Journal of Biol. Med., 39, 289-296 (1967).  Hollingsworth J W , Siegel E R and Creasey W A :  Holton R A , Samoza A R , K i m H B , Liang F, Beidiger R J , Boatman P D , Shindo M , Smith C C , K i m S, Nadizadh H , Suzuki Y , Tao C , V u P, Tang S, Zhang P, Murthi  K K , Gentile L N and L i u JH: First total synthesis of taxol, Journal of American Chem. Soc. 116, 1579-1600 (1994).  Approaches to the total synthesis of taxol, National Cancer Institute Workshop on Taxol and Taxus, National Cancer Institute, Bethesda, MD, 1990.  Holton R A :  Taxol: a new probe for studying the structure and function of microtubules, Cold Spring Harb. Symp. Quant. Biol, 46, 219-226 (1982).  Horwitz S B , Parness J , Schiff P B and Manfredi JJ:  Horwitz SB: Mechanism of action of taxol, Trends Pharmacol. Sci., 13.(4), 134-136 (1992).  M c C u l l o c h C A G and Cruz T F : Paclitaxel selectively induced mitotic arrest and apoptosis in proliferating bovine synoviocytes, Arthritis Rheumatism, 40(6). 1073-1084 (1997).  Hui A , Kulkarni G V , Hunter W L  132  Hunneyball IM: Intra-articular Administration of Drugs, Pharmacy 5, 118-22 (1986).  International,  Ilium L and Davis SS: The targeting of drugs parenterally by use of microspheres, Journal of Parental Science and Technology, 36(6). 242-248 (1982).  Ishida H : Fourier transform infrared characterization of polymers, polymer Sciences and Technology, Vol.36, Plenum Press, New York, (1987).  Itoh K: Intra-articular injection therapy in patient with rheumatoid arthritis, Nippon Rinsho, 50(3), 531-535 (1992).  Iwata M and McGinity JW: Dissolution, solubility, and morphological properties of conventional and multiphase poly(DZ-lactic-co-glycolic acid) microspheres containing water-soluble compounds, Pharmaceutical Research, 10, 1219-1227 (1993).  Jamis-Dow C A , Klecker RW, Sarosy G , Reed E and Collins JM: Steady-state  plasma concentrations and effects of taxol for a 250 mg/m dose in combination with granulocyte-colony stimulating factor in patients with 2  ovarian cancer, Cancer Chemther. Pharmacol,  33, 48-52 (1993).  Jeffery H , Davis SS and O'Hagan DT: The preparation and characterization of  poly(lactide-co-glycolide) microparticles. I: Oil-in-water emulsion solvent evaporation, International Journal of Pharmaceutics,  Jenkins A D : Polymer science, North-Holland,  77_, 169-175 (1991).  Amsterdam, (1972).  Keith HD, Padden FJ Jr. and Vadimsky RG: Intercrystalline links in polyethylene crystallized from the melts, Journal ofPolym. Sci., A-2, 4, 267-281 (1966).  Keller A: A note on single crystals in polymers: Evidence for a folded chain configuration, Philosophical  Magazine, 2, 1171-1175 (1957).  133  Kingston DGI, Hawkins D R and Ovington L: New taxanes from taxus brevifolia, Journal of Natural Products, 45, 466-470 (1982).  Kingston DGI: The chemistry of taxol, Pharmacol. Ther., 52(1), 1-34 (1991).  Kou JH, Emmett C, Shen P, Aswani S, Iwamoto T, Vaghefi F, Cain G and Sanders L:  Bioerosion and biocompatibility of poly(</,l-lactic-co-glycolic acid) implants in brain, Journal of Controlled Release, 43, 123-130 (1997).  Lafyatis, R, Remmers, E F , Roberts A B , Yocum D E , Sporn M B and Wilder RL:  Anchorage-independent growth of synoviocytes from arthritic and normal joints: Stimulation by exogenous platelet-derived growth factor and inhibition by transforming growth factor-beta and retinoids, Journal of Clin. Invest. 83, 1267-1276 (1989).  Lataste H, Senilh V , Wright M , Guenard D and Potier P: Relationships between the  structure of taxol and baccatin III derivatives and their in vitro action on the disassembly of mammalian brain and physarum amoebal microtubules, Proc. Natl. Acad. Sci. USA, 81,4090-4094 (1984).  Lewis D H : Controlled release of bioactive agents from lactide/glycolide polymers, in Chasin M and Langer R (Eds): Biodegrdable polymers as drug delivery systems- drugs and pharmaceutical systems, Vol.45, John Wisley & Son, 1-41 (1990).  Li SM, Garreau H and Vert M : Structure-property relationships in the case of the degradation of massive aliphatic poly-(rx-hydroxy acids) in aqueous media. Part 2: Degradation of lactide-glycolide copolymers: PLA37.5GA25 and PLG75GA25, Journal of materials Science Materials in Medicine ]_, 131-139 (1990B).  Li SM, Garreau H and Vert M : Structure-property relationships in the case of the degradation of massive aliphatic poly-(a-hydroxy acids) in aqueous media. Part 1: Poly(Z)L-lactic acid), Journal of materials Science Materials in Medicine I 123-130 (1990A).  Liggins RT, Hunter W L and Burt H M : Solid-state characterization of paclitaxel, Journal of Pharmaceutical Sciences, 86(12), 1458-1463 (1997).  134  Lin  S Y  and  Pemg  Solid-state interaction studies of drugs/polymers. I.  RI:  Indomethacin/eudragit E , R L . or S resins, STP Pharm. Scl, I 465-471 (1993).  L i n S Y , L e e C J a n d L i n Y Y : Drug-polymer interaction affecting the mechanical properties, adhesion strength and release kinetics of piroxicam-loaded eudragit E films plasticized with different plasticizers, Journal of Controlled Release, 33, 375-381 (1995).  Cell kill kinetics and cell cycle effects of taxol on human and hamster ovarian cell lines, Cancer Chemother.  L o p e s N M , A d a m s E G , Pitts T W and B h u y a n B K :  Pharmacol. 32(3), 235-2^42 (1993).  L y t h g o e B:  The taxus alkaloids. I n  M a n s k e R H F (Eds):  The alkaloids Chemistry  and Physiology, Academic Press, NY, Vol.X, 597-626 (1968).  Magri  N F  and  Kingston  DGI:  Modified taxols. 2. Oxidation products of taxol,  Journal of organic Chemistry, 57, 797-802 (1986).  Modified taxols. 4. Synthesis and biological activity of taxols modified in the side chain, Journal of Natural Products, 57, 298-306  M a g r i N F and Kingston D G I :  (1988).  M a n f r e d i JJ and H o r w i t z S B :  Taxol: an antimitotic agent with a new mechanism  of action, Pharmacol. Ther., 25(1), 83-125 (1984).  Crystal and molecular structure of paclitaxel (taxol), Proc. National Acad. Sci. USA, 92,  Mastropaolo D, Camerman  A, Luo  Y G , Brayer G D  and Camerman  N:  6920-6924 (1995).  A E , M e j i l l a n o M R , N a t h J P , H i m e s R H a n d S t e l l a V J : Synthesis and evaluation of some water-soluble prodrogs and derivatives of taxol and  Mathew  antitumor activity, Journal of Med. Chem., 35(1), 145-151 (1992).  Maulding  HV:  Prolonged delivery of peptides by microcapsules, Journal  Controlled Release, 6,  of  167-176(1987).  135  Mehta  R C , Jeyanthi  R,  Calis  S, Thanoo  B C , Burton K W  and Deluca PP:  Biodegradable microspheres as depot system for parenteral delivery of peptide drugs, Journal of Controlled Release, 29, 375-384 (1994).  Fourier transform infrared spectroscopic study of ion intramolecular interactions in the polar head of  Menikh A and Fragata M :  binding  and  digalactosyldiacylglycerol, Eur. Biophys. J. 22, 249-258 (1993).  Milas L , Hunter N R , Kurdoglu B , Mason K A , M e y n R E , Stephens L C and Peters L J :  Kinetics of mitotic arrest and apoptosis in murine mammary and ovarian tumors treated with taxol, Cancer Chemother. Pharmacol,  35, 297-303 (1995).  Degradation rates of oral resorbable implants (polylactates and polyglycolates): Rate modification with changes in  Miller R A , Brady J M and CutRight D E :  P L A / P G A copolymer ratios, J. Biomed. Mater. Res., IL  711-719 (1977).  Antileukemic alkaloids from taxus wallichiana zucc, Journal of Organic Chemistry, 46, 1469-  Miller R W , Powell R G , Smith C R Jr., Arnold R and Clardy J:  1474 (1981).  Structure and function of articular cartrilage and meniscus. In M o w et al. (Eds): Basic Orthopaedic Biomechanics, Raven  M o w V C , Zhu W B and Ratcliffe A :  Press, New York, 143-198 (1991).  Microencapsulation by coacervation of poly(lactide-co-glycolide) I V . Effect of the processing parameters on coacervation and encapsulation, Journal of Controlled Release,  Nihant  35,  N,  Grandfils  C,  Jerome  R  and  Teyssie P:  117-125(1995).  Nocolaou K C , Yang Z , L i u JJ, Ueno H , Nantermet P G , G u y R K , Claiborne C F , Renaud J , Couladouros E A , Paulvannan K and Sorensen E J :  Total synthesis of  taxol, Nature, 367(2), 630-634 (1994).  O'Shaughnessy J A and Cowan K H :  Current status of paclitaxel in the treatment  of breast cancer, Breast Cancer Research and Treatment, 33, 27-37 (1994).  136  The preparation and characterization of poly(lactide-co-glycolide) microparticles: III. Microparticle/polymer degradation rates and the in vitro release of a model protein, International  O'Hagan D T , Jeffery H and Davis SS:  Journal of Pharmaceutics,  103. 37-45 (1994).  Ojima I, Habus I, Zhao M , Georg GI, and Jayasinghe L R : Efficient and practical  asymmetric synthesis of the taxol C-13 side chain, 7V-benzoyl-(2i?, 35)-3phenylisoserine and its analogues via chiral 3-hydroxy-4-aryl-/?-lactams through chrial ester enolate-imine cyclocondensation, Journal of Organic Chemistry, 56, 1681-1683 (1991).  Okada H and Toguchi H :  Biodegradable microspheres in drug delivery, Critical  Review in Therapeutic Drug Carrier Systems, I2_(l), 1-99 (1995).  Suppression of collagen-induced arthritis using an angiogenesis inhibitor, AGM-1470, and a microtubule stabilizer,  Oliver SJ, Banquerigo M L and Brahn E :  taxol, Cellular Immunology, I5Z 291-299 (1994).  Onetto N , Canetta R, Winograd B , Catane R, Dougan M , Grechko J , Burroughs J and Rozencweig M :  Overview of taxol safety, Journal of the National  Cancer  Institute Monographs, L5, 131-139 (1993).  Disappearance kinetics of solutes from intra-articular injection, British Journal of Clin.  Owen S G , Francis H W and Roberts M S :  synovial Pharmacol,  fluid  after  38(4), 349-355 (1994).  Highly stereo-selective synthesis of a-hydroxy p-amino acids through |3-lactams: application to the synthesis of the taxol and bestatin side chains and related  Palomo C , Arrieta A , Cossip F, Aizpurua J M , Mielgo A and Aurrekoetxea N :  systems, Tetrahedron Letter, 3£ 6429-6432 (1990).  Park T G :  Degradation of poly(Iactic-co-glycolic acid) microspheres: effect of  copolymer composition, Biomaterials,  Park H and Park K:  Pharmaceutical  16(15). 1123-1130 (1995).  Biocompatibility issues of implantable drug delivery systems, Research, 13(2). 1770-1776 (1996).  137  Taxol binds to polymerized tubulin in vitro, Journal of Cell Biology, 91 479-487 (1981).  Pamess J and Horwitz S B :  Perrone R K and Stenberg SR:  Crystalline paclitaxel hydrates, EP 0 717 041 Al,  (1996).  T A and Schindler A : Biodegradable drug delivery systems based on aliphatic polyesters: Application to contraceptives and narcotic antagonists, NIDA Res. Monogr., 28, 232-53 (1981).  Pitt C G , Marks  Glass transition temperature in hydrogen-bonded polymer mixtures, Polymer, 36(6), 1223-1227 (1995).  Prinos J and Panayiotou C :  Experimental methods in polymer chemistry - physical principles and applications, John Wiley & Sons, (1980).  Rabek IF:  The influence of manufacturing procedure on the degradation of poly(lactide-co-glycolide) 85:15 and 50:50 implants, Journal of Controlled Release, 43, 161-173 (1997).  Ramchandani M , Pankaskie M and Robinson D:  A and Cox C : Taxol and taxotere in bladder cancer: In vitro activity and urine stability, Cancer Chemther. Pharmacol, 33, 460-464 (1994).  Rangel C , Niell H , Miller  A , Wilson C G and Davis SS: Preparation and evaluation of biodegradable polymeric systems for the intra-articular delivery of drug, Journal of Pharmacy and Pharmacology, 36, 431-436 (1984A).  Ratcliffe J H , Hunrieyball IM, Smith  Wilson C G , Smith A and Davis SS: Albumin microspheres for intra-articular drug delivery: Investigation of their retention in normal and arthritic knee joints of rabbits, Journal of Pharmacy and Pharmacology, 39, 290-295 (1987).  Ratcliffe J H , Hunneyball IM,  Ratcliffe J H , Hunneyball IM, Wilson C G , Smith A and Davis SS: The potential of  albumin microspheres as an intra-articular drug delivery system, British Journal of Rheumatoloy, 23, 118 (1984B).  138  Biodegradable polymers for use in surgery poly(glycolic)/poly(lactic acid) homo and copolymers: 2. In vitro degradation,  Reed A M  and Gilding D K :  Polymer, 22, 494-498 (1981).  Ringel I and Horwitz S B :  Studies with RP56976 (taxotere): A semisynthetic  analogue of taxol, Journal of National Cencer Inst., 83, 288-291 (1991).  Taxol is converted to 7-epitaxol, a biologically active isomer, in cell culture medium, Journal of Pharmac. Exp. Ther., 242. 692-698  Ringel I and Horwitz S B :  (1987).  Roeges N P G : A  guide to the complete interpretation of infrared spectra of  organic structures, John Wiley & Sons, (1994).  i Rowinsky E K , Cazenave L A and Donehower R C :  Taxol: a novel investigational  antimicotubule agent, Journal of National Cancer Institute, 82(15). (1990).  1247-1259  Rowinsky E K , Eisenhauer E A , Chaudhry V , Arbuck S G and Donehower R C :  Clinical toxicities encountered with paclitaxel (taxol®), Seminars in Oncology, 20(4), Suppl 3, 1-15 (1993).  Rowinsky E K , Onetto N , Canetta R M and Arbuck, S G : Paclitaxel:  the first of the taxanes, an important new class of antitumor agents, Semin. Oncol., 19, 646662 (1992).  In vivo peptide release from poly(2)£-lactic acid-coglycolic acid) copolymer 50/50 microspheres, Journal of Controlled Release,  Ruiz J M and Benoit JP:  16,  177-186(1991).  Influence of average molecular weights of poly(Z)L-lactic acid-co-glycolic acid) copolymers 50/50 on phase separation and in vitro drug release from microspheres, Pharmaceutical Research, 7(9),  Ruiz J M , Busnel JP and Benoit JP:  928-934 (1990).  Prolonged controlled-release of nafarelin, a luteinizing hormone-releasing hormone analogue, from  Sanders L M , K e l l B A , M c R a e GI and Whitehead G W :  139  biodegradable polymeric implants: Influence of composition and molecular weight of polymer, Journal of Pharmaceutical Sciences, 75(4), 356-60 (1986).  Schakenraad J M , Nieuwenhuis P, Molenaar I, Helder J, Dijkstra PJ and Feijen J: In  vitro and in vivo degradation of glycine/rf,/-lactic acid copolymers, 23, 12711288 (1989).  Schiff PB and Horwitz SB: Taxol stabilizes microtubules in mouse fibroblast cells, Proc. Natl. Acad. Sci. USA, 77(3), 1561-1565 (1980).  Schiff P B , Fant J and Horwitz S B :  Promotion of microtubule assembly in vitro by  taxol, Nature, 277(5698), 665-667 (1979).  See Y : Intra-synovial corticosteroid injections in juvenile chronic arthritis - a review, Annals Academy of Medicine, 27(1). 105-111 (1998).  Poly(glycolic acid-co-rf,/-lactic acid) : Diffusion or degradation controlled drug delivery ?, Journal of Controlled Release, 18, 261-  Shah SS, C h a Y and Pitt C G :  270 (1992).  Shah N H , Railkar A S , Chen F C , Tarantino R, Kumar S, Murjani M , Palmer D , Infeld  A biodegradable injectable implant for delivering micro and macromolecules using poly(lactic-co-glycolic) acid (PLGA) copolymers,  M H and Malick A W :  Journal of Controlled Release, 27, 139-147 (1993).  Liposomeincorporated corticisteroids: I. The interaction of liposomeal Cortisol palmitate with inflammatory synovial membrane, British Journal of Exp.  Shaw  IH,  Knight  C G , Thomas D P , Philips N C and Dingle  JT:  Path., 60(2), 142-150 (1979).  Spenlehauer G , Vert M , Benoit JP and Boddaert A :  of  poly(Z)£-lactide-co-glycolide) type  In vitro and in vivo degradation  microspheres  made  by solvent  evaporation method, Biomaterials, 10, 557-563 (1989).  Biodegradable cisplatin microspheres prepared by the solvent evaporation method:  Spenlehauer G , Vert M , Benoit JP, Chabot F and Veillard M :  140  Morphology and release characteristics, Journal of Controlled Release, 7, 217229 (1988).  Sperling L H :  Introduction to physical polymer science, John Wiley & Sons, (1986).  Preparation of biodegradable poly(lactic-co-glycolic) acid microspheres and their in vitro release of timolol  Sturesson C , Carlfors J, Edsman K and Andersson M :  maleate, International Journal of Pharmaceutics,  Suffness M :  82, 235-244  (1993).  Is taxol a surrogate for a universal regulator of mitosis? In Vivo, 8(5),  867-878 (1994A).  Overview of paclitaxel research - progress on many fronts. In Georg GI et al. (Eds): Taxane anticancer agents -basic science and current status,  Suffness M :  ACS symposium Series 583, Library of Congress, 1-17 (1994B).  Biologically active taxol analogues with deleted A-ring side chain substituents and variable C-2'  Swindell C S , Krauss N E , Horwitz S B and Ringel I:  configurations, Journal of Med. Chem. 34, 1176-1184  (1991).  Preparation and characterization of copoly(</,/-Iactic/glycolic acid) microparticles for sustained release of thyrotropin releasing hormone by double nozzle spray drying method,  Takada S, U d a Y , Toguchi H and Ogawa Y :  Journal of Controlled Release, 32, 79-85 (1994).  Tamada J and Langer R:  Erosion mechanism of hydrolytically degradable  polymers, Proc. Nat. Acad. Sci. USA, 90, 552-556 (1993).  A new parenteral vehicle for the administration of some poorly water-soluble anti-cancer drugs, Journal of Parenteral Sci.  Tarr B D and Yalkowsky S H :  Technol, 41, 31-33 (1987).  The Merck Index - an encyclopedia of chemicals, drugs, and biologicals, 12 Ed., pp.1200, (1996).  141  In vivo degradation of massive poly(alpha-hydroxy acids): Validation of in vitro findings, Biomaterials, 13(9). 594-600 (1992).  Therin M , Christel P, L i S, Garreau H and Vert M :  Injectable, long-acting microparticle formulation for the delivery of anti-inflammatory agents, U.S. Patent 4,542,025,(1985).  Tice  TR,  Turi A :  Lewis  DH,  Cowsar  DR  and  Beck  LR:  Thermal characterization of polymeric materials, Academic Press, (1997).  More about the degradation of LA/GA-derived matrices in aqueous media, Journal of Controlled Release, 16, 15-26 (1991).  Vert M , L i S and Garreau H :  Visscher G E , Pearson J E , Fong J W , Argentieri G J , Robison R L and Maulding H V :  Effect of particle size on the in vitro and in vivo degradation rates of poly(</,/lactide-co-glycolide) microcapsules, Journal of Biomedical Materials Research, 22, 733-746(1988).  Visscher G E , Robison R L , Maulding H V , Fong J W , Pearson J E and Argentieri G J :  Biodegradation of and tissue reaction to 50:50 poly(rf,Mactide-co-glycolide) microcapsules, Journal of Biomedical Materials Research, 19, 349-365 (1985).  Volland C , Wolff M and Kissel T: The influence of terminal gamma-sterilization  on captopril containing poly(</,/-lactide-co-glycolide) microspheres, Journal of Controlled Release, 3£ 293-305 (1994).  Influence of physicochemical properties of polylactic acid on the characteristics and in vivo release patterns of polylactic acid microspheres containing local anesthetics, Chem. Pharm. Bull, 30(7). 2621-2628 (1982).  Wakiyama N , Juni K and Nakano M :  Plant antitumor agents. V I . The isolation and structure of taxol, a novel antileukemic and antitumor agent from Taxus brevifolia, Journal of the American Chemical Society, 93(9), 2325-2327 (1971).  Wani M C , Taylor H L , Wall M E , Coggon P and McPhail A T :  142  Weiss  S, Kisch  E S and Fischel  B : Systemic  effects  of  intra-articular  administration of triamcinolone hexacetonide, Isr. J. Med. Sci. 19.(1), 83-84 (1983).  Winternitz CI (unpublished results):  Solubility of paclitaxel in phosphate buffer  saline with albumin, (1997).  Thermal characteristics of poly(DL-lactic acid) microspheres containing neurotensin  Yamakawa I, Ashizawa K , Tsuda T , Watanabe S, Hayashi M and Awazu S:  analogue, Chem. Pharm Bull., 40, 2870-2875 (1992).  Y e h G S Y : Order in amorphous polystyrene as revealed by electron diffraction and diffraction microscope, Journal of Macromol. Sci. Phys., 6(3), 451-463 (1972).  Yoshioka S, A s o Y , Otsuka T and Kojima S:  The effect of y-irradiation on drug  release from poly(lactide) microspheres, Radial. Phys. Chem., 46(2), 281-285 (1995).  Mechanism of lactide polymerization in the presence of stannous octoate: The effect of hydroxy  Zhang X C , Macdonald D A , Goosen M F A and Mcauley K B :  and carboxylic acid substances, Journal of Polymer Science: Part A: Polymer Chemistry, 32, 2965-2970 (1994).  Zhang X C (unpublished results):  Solubility of paclitaxel in PBS-albumin, (1997).  Zoli W , Flamigni A , Frassineti G L , Bajorko P, D e Paola F, Milandri C , Amadori D  activity of taxol and taxotere in comparision with doxorubicin and cisplatin on primary cell cultures of human breast and Gasperi-Campani A : In vitro  cancers, Breast Cancer Research and Treatment, 34, 63-69 (1995).  143  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0089168/manifest

Comment

Related Items