Influence of physico-chemical parameters on the release of hydrocortisone acetate from albumin microspheres

E L S E V I E R Journal of Controlled Release 31 (1994) 61-71

journal of controlled

re l ease

Influence of physico-chemical parameters on the release of hydrocortisone acetate from albumin microspheres

Isabella Orienti*, Antonietta Coppola, Elisabetta Gianasi, Vittorio Zecchi Dipartimento di Scienze Farmaceutiche, Via S. Donato 19, 40127 Bologna, Italy

Received 24 August 1993; accepted in revised form 6 January 1994

Abstract

We describe the preparation of bovine serum albumin microparticles, loaded with hydrocortisone acetate by mixing different drug dispersing phases with the protein aqueous solution before thermal crosslinking. The influence of the drug dispersing phase is analysed in terms of the physico-chemical properties correlated to release. The release of the drug from the microparticles is initially Fickian or Anomalous, subsequently approaches zero order and finally is exponentially correlated to time. The drug dispersing phases able to induce hydrophilic modifications on the polymeric structure, increase the duration of the zero order release owing to modifications on the diffusion coefficient and solubility of the drug in the matrix.

Keywords: Bovine serum albumin; Microparticle; Hydrocortisone acetate; Drug dispersing phase; Thermal crosslinking; Drug loading; Diffusion; Solubility; Bulk erosion; Release kinetics

1. Introduction

The utilization of albumin in micro- and nanoparti- culate preparations has recently received increasing attention particularly in passive drug targeting [ 1--4]. Albumin is suitable for producing non antigenic nano- and microparticulates whose physico-chemical prop- erties can be widely modulated according to the crosslinking method employed for their production and the nature of the drug [5-12]. The drug may be released from the particles according to kinetics approaching zero order for defined periods of time [6,7,13 ]. Albumin microparticles can be obtained by chemical or thermal crosslinking of the albumin mac- romolecules solubilized in the aqueous phase of a w/o emulsion [ 14]. Lipophilic drugs may be loaded in the albumin microparticles by incorporation of a drug dis-

*Corresponding author.

0168-3659/94/$07.00 © 1994 Elsevier Science B.V. All rights reserved SSDI0168-3659 ( 9 4 ) 0 0 0 0 4 - E

persing phase in the albumin containing aqueous phase of the w/o emulsion. The drug dispersing phase is chosen to favour the presence of the lipophilic drug in the albumin containing aqueous phase during cross- linking, allowing good loading levels in the micropar- ticles [ 13 ].

In this work we describe the preparation of bovine serum albumin (BSA) microparticles loaded with hydrocortisone acetate in the presence of different drug dispersing phases: corn oil, isopropyl alcohol and n- butyl alcohol. By different mechanisms, these phases modify the drug distribution, both solubilized or in the solid state, between the two phases of the w/o emulsion and therefore the loading levels of the microparticles. The influence of the drug dispersing phases on the variation of the drug diffusion coefficient and solubility in the polymeric structure over time was also investi- gated. Lastly, the effect of a viscosity enhancing agent

62 I. Orienti et ~d. /Jaurnal qf Controlled Release 31 (1994) 61 71

(polyacrylic acid) on the microparticle properties and release was analysed.

with a mean geometric diameter of 20_ 5 #m were selected for the present study.

2. Experimental

2.1. Materials

BSA, gelatin type B, corn oil (CO) and hydrocor- tisone acetate (HCA) were all purchased from Sigma Chemicals (St. Louis, MO, USA). Isopropyl alcohol (ISP), n-butyl alcohol (nBU) and diethyl ether were purchased from Carlo Erba Analyticals (Milan, Italy). Polyacrylic acid (Carbopo1940, CBP) was kindly sup- plied by Biochim (Milan, Italy).

2.2. Microparticles preparation

Albumin microparticles were prepared by dispersing 50 mg of hydrocortisone acetate in 0.5 ml isopropyl alcohol, n-butyl alcohol or corn oil, respectively. The dispersion was mixed with 1 ml of an aqueous solution containing 25% BSA and 1% gelatin used to enhance the mixing of the drug dispersing phase with the aque- ous phase [ 15,16]. The mixture, stirred for 10 min at 25°C, was subsequently frozen to 0°C and added to 6 ml corn oil at the same temperature stirring for 2 rain before adding to 100 ml of corn oil heated to 120°C. Thermal crosslinking was carried out for 20 min in a glass beaker under vigorous stirring (750 rpm) using a four-bladed impeller (4 cm diameter). The CBP con- taining microparticles were prepared by adding 2.5 mg CBP to the drug dispersing phase. The microparticles obtained were separated by centrifugation and washed with 100 ml diethyl ether. The particles were subse- quently warmed to 40°C under a vacuum to a constant weight to evaporate the isopropyl alcohol and n-butyl alcohol and form an internal completely solid phase which suggests microspheres formation. In the pres- ence of corn oil as dispersing phase the permanence of a liquid phase within the particles suggests the forma- tion of microcapsules. This observation was further confirmed by optical microscopic studies of the micro- particles (Axiophot-Zeiss, Germany). The particles were sieved (Giuliani certified test sieves, Turin, Italy). The dimensions of the particles ranged from a minimum of 5 to a maximum of 50/zm. Microparticles

2.3. Viscosi~ tests

The viscosities of the aqueous phases, utilized tor the preparation of the microparticles containing BSA and gelatin with or without CBP, were tested by a Rheotest 2 MLW Freital, DDR viscosimeter.

2.4. Determination of drug content in the microparticles

The amount of hydrocortisone acetate entrapped in the microparticles was determined by digesting the microparticles in a mixture of ethanol and NaOH 0.5 N aqueous solution and spectrophotometricaily detect- ing the drug in the mixture with respect to that obtained by digesting the unloaded microparticles in the same conditions.

2.5. X-Ray microspheres analysis

X-Ray diffraction analysis of the microparticles was performed using Philips PN 1130 powder diffracto- meter Cu-K,~ radiation, l°/min scan speed.

2.6. Swelling studies

Swelling was followed by microscopic observation (Axiophot-Zeiss, Germany) of twenty microparticles of each type immersed in water at 37°C. The increase in diameter was followed as long as the spherical shape was maintained.

2.7. Determination of HCA solubility in the microparticles over time

Unloaded microparticles, prepared by the same method as loaded microparticles, were soaked in aque- ous HCA saturated solutions for different periods of time at 37°C to determine the solubility of HCA (C~) after defined periods. Solubility was obtained by deter- mination of the drug content in the microparticles after the appropriate period of soaking.

L Orienti et al./Journal of Controlled Release 31 (1994) 61-71 63

2.8. Diffusion coefficient studies

The diffusion coefficients of HCA in the micropar- ticles were measured by the release profiles of the unloaded microparticles, prepared by the same method as loaded microparticles, soaked with aqueous HCA saturated solutions for different periods of time at 37 °C. The diffusion coefficients characteristic of the dif- ferent periods were calculated from the slopes of the release curves [17] relative to the various periods according to the equation:

D= 7rr2 /36[ d(MJM~) /dt 1/2] 2 up to Mt/M~ <_0.

(1)

where r is the radius of the soaked microspheres.

2.9. In vitro release studies

20 mg of loaded or unloaded microparticles were suspended in 5 ml of water. The suspension was placed in a donor cell separated by a dialysis membrane from a receiving compartment containing 10 ml of water which was replaced after time intervals suitable to guar- antee sink conditions throughout the runs [18]. The system was thermostated at 37°C. The drug was spec- trophotometrically detected in the receiving phase over time, utilizing the unloaded microparticles as a blank. In order to avoid microbial contamination, the water was filtered through a millipore filter of 0.45/xm pore size and the instrumentation was sterilized in a pressure cooker at 121°C for 21 min. At the end of the release studies microbiological tests were performed indicat- ing that the liquid was not contaminated.

3. Results

3.1. Drug loading

Table 1 HCA content (%, w:w) 5: SD, of BSA and BSA/CBP microparticles loaded by different drug dispersing phases mixed with the BSA aqueous solution before crosslinking: corn oil (CO), isopropyl alco- hol ( ISP ) and n-butyl alcohol (nBU) a

Dispersing BSA BSA/CBP phases

CO 9.31 + 0.89 9.85 5:0.97 1SP 6.98 5:1.03 7.99 5:1.2 I nBU 10.27 5:0.95 10.37 + 1.40

aMeans of three batches of microparticles.

phase (CO, nBu), increased affinity of the BSA con- taining aqueous phase for the drug induced by the pres- ence of a water miscible drug dispersing phase (ISP). The presence of a viscosity enhancing agent such as CBP in the BSA containing aqueous phase may be a further element hindering drug migration towards the lipidic phase (viscosity of the 25% BSA and 1% gelatin aqueous phase without CBP: 826.04 cps; with CBP: 1998.57 cps). Table 1 reports the loading levels of the BSA microparticles. The microparticles obtained in the presence of n-butyl alcohol (M-nBU) are character- ized by higher loading levels than those obtained in the presence of corn oil (M-CO) and isopropyl alcohol, respectively (M-ISP). The presence of CBP slightly increased the loading levels. These data seem to outline the prevalent effect of the immiscibility of the drug dispersing phase with the BSA containing aqueous phase in the loading process and the favourable effect produced by an increase in viscosity of the aqueous phase. Moreover the higher loading obtained with nBu than CO may be attributed to the low miscibility of nBu with the external lipidic phase (CO) of the w/o emul- sion which further hinders drug migration.

3.2. X-Ray diffraction studies

The loading of the BSA microparticles here described may be prevalently attributed to the ability of the drug dispersing phase to hinder drug migration either in solubilized or solid form from the aqueous BSA containing phase towards the external lipidic phase (CO) of the w/o emulsion. The hindering effect may be a consequence of: immiscibility of the drug dispersing phase with the BSA containing aqueous

The X-Ray diffraction scans show that in all the microparticles studied the drug is present in the solid state, both crystalline and amorphous. Fig. 1 shows that the crystallinity is higher in M-CO and decreases in M- ISP and M-nBu respectively. No significant differences were observed in the microparticles containing CBP with respect to the corresponding microparticles with- out CBP.

64 1. Orienti et al. /Journal of Controlled Release 31 (1994) 61-71

4[~))

3: ' (y

3(~XY

2506-

2 t ~

1000-

500-

(A)

10 15 20 25 3o 35 40 45 50

2@

,BJ

0 5 10 15 20 25 30 35 40 45

2 0

(C)

45O-

4CCr

350 ~

i 2 0

(D)

2~ 2

' i

~55 10 15 20 25 30 35 40 45

2 o

Fig. 1. X-Ray diffractiometric analysis of: (A) crystalline hydrocortisone acetate; (B) microparticles loaded using corn oil as drug dispersing phase; (C) microparticles loaded using isopropyl alcohol as drug dispersing phase; (D) microparticles loaded using n-butyl alcohol as drug dispersing phase.

(A) (B)

2,0

o

L5

1,0"

0 Z

0,5

2.0

o

~, ~,5

"~ 1,0

0

Z 0,5

E ' " f3---~

50 100 150 200 250 50 100 150 200 250

Time (h) Time (h)

Fig. 2. Swelling of: (A) BSA microparticles prepared in the presence of: ( • ) corn oil, ( 1 ) isopropyl alcohol, ( e ) n-butyl alcohol, in water at 37°C; and (B) BSA/CBP microparticles prepared in the presence of: (~ ) corn oil, ( [ ] ) isopropyl alcohol (O) n-butyl alcohol, in water at 37°C.

1. Orienti et al. /Journal of Controlled Release 31 (1994) 61-71 65

(A) (B)

-20-

- 2 2

- 2 4

-26

-28

-30

-20"

- 2 2

-24-

"-~ - 2 6

- 2 8

. . . . . -30 50 100 150 200 250

Time (h)

50 100 150 200 250

Time (h)

Fig. 3. Diffusion coefficient of HCA in: (A) BSA microparticles, prepared in the presence of: ( • ) corn oil, (11) isopropyl alcohol, (o) n- butyl alcohol, undergoing bulk erosion; and (B) BSA/CBP microparticles, prepared in the presence of: ( A ) corn oil, (I-q) isopropyl alcohol, (O) n-butyl alcohol, undergoing bulk erosion.

3.3. Swelling studies

Fig. 2 reports the normalized diameter increase vs time (dr~do) of the microparticles in aqueous phase. In each system an initial increase was followed by a decrease to a plateau. Fig. 2 also shows a more enhanced increase in M-ISP than M-nBU and M-CO respectively, the presence of CBP further enhances the normalized diameter increase. The effect of the differ- ent drug dispersing phases on the hydrophilic character of the polymeric structure may be attributed to their ability to interact with the polymeric chain partially modifying its conformation making it more unfolded, able to form more hydrogen bonds with the water mol- ecules [ 19,20]. This effect is expected to be more pro- nounced in the presence of the water miscible ISP alcohol than the unmiscible nBU alcohol while it is not expected in the presence of CO.

3.4. Time dependence of HCA diffusion coefficient in the microparticles

In the systems undergoing a bulk erosion a depend- ence of the drug diffusion coefficient on time according to the equation [21-23]:

D = Do ekot (2)

has been proved where Do is the diffusion coefficient at t = 0 and kD is the erosion constant characteristic of the drug-releasing system. According to the logarith-

mic form of Eq. 2. Fig. 3 shows that for each system studied In D linearly increases with time following two different trends, the first (first D increasing period) having a higher slope than the second (second D increasing period). The linear correlations obtained indicate that in each system studied a bulk erosion takes place which exponentially increases the drug diffusion coefficient over time in accordance with two different erosion constant values, the first being higher than the second. The erosion constant values determined from the slopes of the first and second linear trends of each system are reported in Table 2 as kD~ and kD2, respec- tively. The decrease in the erosion constant value sug- gests that a partial rearrangement of the polymeric backbone undergoing erosion may take place at hydrol- ysis levels characteristic of each system with the for- mation of structures less available to hydrogen bound with water molecules and consequently less available to hydrolysis. This hypothesis is further supported by the observed decrease in the normalized diameter of the microparticles which starts, for each system, during its second D increasing period (Figs. 2 and 3). The partial rearrangement of the polymeric backbone towards less hydrophilic structures probably triggers a compactation of the polymeric matrix accounting for the normalized diameter decrease. Table 2 and Fig. 3 show that kDl and kD2 values follow the sequence:

M-ISP/CBP > M-ISP > M-nBU/CBP

> M-nBU > M-CO/CBP > M-CO

66 I. Orient± et al. / Journal oJ Controlled Release 31 (1994) 61-71

Table 2a Physico-chemical parameters of HCA loaded BSA and BSA/CBP microparticles at 37°C in water: diffusion coefficient of HCA in the microparticles (D) (means of three release studies)

Dispersing First D increasing period Second D increasing period phases D - D,, e ~''' D = Do2 e kt'''

Dol ± SD kDi ± SD Do2 ± SD ki) 2 ± SI) (cm2/s×10 '~) (h I×10-~) (cm2/sXl0 -Is) ( h " × 1 0 ~)

BSA microparticles CO 22.06_+2.65 10.01 ± 1.18 24.10+3.34 9.28± 1.13 IS P 0.54 ± 0.04 50.91 ± 4.59 3.88 ± 0.29 28.09 + 2.85 nB U 3.39 ± 0.09 31.86 ± 1.11 13.25 -+ 0.56 18.97 ± 1.04

BSA/CBP microparticles CO 3.79 ± 0.56 31.20 ± 3.87 12.20 ± 1.97 18.80 ± 2.37 ISP 0.13±0.01 76.48±8.32 1.47±0.15 45.46±5.07 nBU 2.50 ± 0.11 40.21 ± 2.69 8.59 ± 0.49 25.97 ± 2.05

Table 2b Solubility of HCA in the microparticles (C~) (means of three batches of microparticles)

Dispeming First Cs decreasing period Second Ca decreasing period Third Cs decreasing period phases Cs=k' e -~'~' C~=l(' e -k,.,., C.~=Ig" e -~:~'

k ' ( m g / m g x l O -~) ~.~(h-l>(10 -3) k" (mg/mg×10 3) kc2(h-lx10-3) k,,,(mg/mgxl0-3) kc3(h-l×10-3) ±SD ±SD ±SD ±SD ±SD ±SD

BSA mieropartieles CO 6.05 ± 1.03 8.33 ± 1.12 4.16 ±0.74 3.75 ±0.53 1.02 ±0.58 1.61 +0.24 ISP 41.58±5.15 54.16±4.41 21.15±3.07 26.00±2.29 2.60±0.39 1.79±0.18 nBU 10.73 ± 0.45 33.75 ± 0.88 4.65 ± 0.26 16.25 ± 0.62 1.09 ± 0.08 1.67 ± 0.07

BSAJCBP mieropartieles CO 8.31 q- 1.53 28.33 _ 4.31 3.76 ± 0.70 13.96 + 2.18 1.07 + 0.20 1.84 ± 0.29 ISP 57.26 ± 8.78 71.25 ± 7.84 33.60 ± 5.45 34.33 ± 3.91 2.70 ± 0.45 2.16 ± 0.27 nBU 12.15 ± 1.08 37.08± 1 . 8 2 6.85±0.63 17.87± 1 . 0 0 1.72±0.19 1.87+0.13

This sequence, being in accordance with that observed for the normalized diameter increase (Fig. 2) , may be explained by an enhanced peptide l inkage hydrolysis due to higher water concentrat ions, in those polymeric structures where the interactions with the drug dispers- ing phase have increased their hydrophil ic character.

3.5. T ime d e p e n d e n c e o f H C A so lub i l i t y in the

m ic ropar t i c l e s

The progressive hydrolysis of the peptidic bonds during erosion is expected to increase the hydrophilic- ity of the polymeric network consequent ly decreasing the solubil i ty of the l ipophilic HCA in the microparti-

cles. Fig. 4 reports the In Cs as a function of t ime and shows that for each system In Cs decreases over t ime according to three linear trends. The first trend is char-

acterized by a higher slope (first Cs decreasing period) than the second (second Cs decreasing period). The third (third Cs decreasing period) is characterized by a m i n i m u m slope whose value approaches this trend to a plateau. The l inear dependence of In Cs on t fits with an exponential correlation between Cs and t:

Cs = ke - k,-t (3)

where k is the intercept and kc is the slope of each l inear trend obtained from In C~ vs t. For each system the slopes of the first, second and third l inear trend are

I. Orienti et al. / Journal of ControUed Release 31 (1994) 61-71 67

(A) (B)

-3" -3

-4 ~ -4

~ -6

.7 ̧

.8 ̧ .8 ̧ 0 50 100 150 200 250 0 50 100 150 200 250

Time (h) Time (h)

Fig. 4. Solubility of HCA in: (A) BSA microparticles, prepared in the presence of: ( • ) corn oil, ( l l ) isopropyl alcohol, (e) n-butyl alcohol, undergoing bulk erosion; and (B) BSA/CBP microparticles, prepared in the presence of: ( ~ ) corn oil, (I-7) isopropyl alcohol, (O) n-butyl alcohol, undergoing bulk erosion.

reported in Table 2 as kc~, kcz and kc3, respectively, and the relative intercepts as k', k", k".

3.6. Comparison between the variations of the diffusion coefficient and solubility of HCA over time

During the second D increasing period, the decrease in the erosion constant (kD2 < kin) and the consequently slowed transformations of the polymeric structure fur- ther slow the C~ decrease, the Cs values being virtually constant over time (kc3 < < kD2).

The comparison between the duration of the D increase and Cs decrease periods (Table 2 and Figs. 3 and 4), as determined by the intersections of straight lines corresponding to the linear trends, show that for each system the first D increasing period corresponds to both the first and second C~ decreasing periods while the second D increasing period corresponds to the third C~ decreasing period. Moreover the kDl values approach kci and 2kc2 for each system, while kD2 is always much higher than kc3.

These correlations suggest that during the first D increasing period the decrease in C~ is controlled by erosion, kD~ being - kcl, until some form of hydro- phobic side chain aggregation starts to operate, due to an enhanced flexibility of the polymeric backbone undergoing hydrolysis, attenuating the rate of hydro- philic transformations of the polymeric structure and consequently attenuating the C~ decrease rate in accor- dance with the decrease in the kc value from kc~ to kc2. When the decrease in the kc value takes place the Cs decrease may be supposed to depend on both the ero- sion process and the transformations induced by ero- sion in the polymeric structure, being the erosion constant value higher than the kc2 value (kDl = 2kc2) .

3.7. Release studies

Kinetic analysis of release From the kinetic analysis of release carried out in

each system according to the general equation [24- 25]:

M,/M~ = kt" (4)

the n values were determined for the different periods which characterize the Cs decrease over time up to M,/ M= < 0.6 (Table 3). The n values obtained in the first Cs decrease period range from 0.37 to 0.62, suggesting a prevalently Fickian or Anomalous release which may be explained according to an almost complete counter- balance of the D increase and Cs decrease due to the similar values of the exponential constants obtained in this period (kin--kcl ). The substitution of Eqs. 2 and 3 in the equation:

M=S( DCs2At) 1/2 (5)

which describes the release from diffusion controlled releasing systems [ 21 ], where S is the surface of the release system and A the drug amount incorporated in the unit volume of the system, leads to the equation:

68 I. Orient± et fd. / Journal qf Cmttrolled Release 31 (1994) 6t-71

M = SI D,, jk 'e I~''' k"~'2At] J/2 (6)

suitable for systems where drug release results from both diffusion and bulk erosion [21]. During the release period where k~ = k o , Eq. (6) reduces to:

M = S( D,,Ik '2At) t/2 (7)

accounting for the Fickian or Anomalous releases obtained from each system in this period. The n values

obtained in the second C~ decrease period range from 0.76 to 0.97, suggesting release kinetics approaching

zero order. This behaviour may be attributed to the changed correlations between the D increase and the

C~ decrease, being ko~ =2kc2 in this period. If the releasing system is controlled by diffusion, the substi-

tution of Eqs. 2 and 3, with the appropriate constant values, into Eq. 5 leads to:

M = S [ Doj k"e ~m - k(2)t2At ] 1/2 (8)

where the square root function of time, accounting for

the release rate decrease due to the increased diffusional

pathway over time [ 17], may be compensated by the exponential function of time, resulting from the prev- alence of D increase on C~ decrease. This situation may therefore account for the n values obtained in this

period if diffusion controls the release. If the releasing

systems have reached sufficiently high D values to be controlled by drug dissolution, the substitution of Eqs. 2 and 3 in the equation:

M = 2SC~( DK) i/2[ ( 1/2K) + t ] (9)

which describes the release from dissolution controlled release systems [26-28] , where K is the dissolution constant of the drug in the releasing system, leads to equation:

M = 2Sk"(Ool K) l/2e ¢ !/2kl)1- I,~:2)t[ ( 1/2K) + t]

( ]0)

In this period/%1 being - 2kc2, Eq. 10 is reduced to:

M = 2 S k " ( D o I K ) t / 2 [ ( 1 / 2 K ) + t ] ( 11 )

Table 3 Release characteristics of BSA and BSA/CBP microparticles: duration of the different release period (A,, h) ± SD related diffusional exponents (n) + SD and drug fraction released (M,/M~) 5: SD (means of three release studies)

Dispersing phases First period Second period Third period

BSA microparticles CO A, 81.60 ± 7.40 38.40 ± 3.65 90.00 ± 9.09

n 0.46 ± 0.04 - MffM,~ 0.57 ± 0.06 0.11 5:0.04 0.32 ± 0.01

A, 24.00 ± 1.46 62.40 ± 4.46 268.60 ± 20.25 n 0.37 ± 0.02 0.76 5:0.05 M,/M~ 0.25 ± 0.03 0.20 ± 0.02 0.55 ± 0.04

A, 48.005:1.01 51.605:1.96 380.405:17.12 n 0.40 ± 0.01 0.81 ±0.04 M,/M~ 0.39 ± 0.02 0.18 ± 0.01 0.43 ± 0.03

ISP

nBU

BSA/CBP microparticles CO A,

n M,/M~

ISP A t

v/

M,/M~

nBU A t

/2

M,/M~

55.20 5:5.63 48.24 ± 5.00 146.60 ± 15.36 0.50 ± 0.05 0.51 ±0.06 0.18±0.03 0.31 ±0.04

14.40 ± 1.12 64.08 ± 5.26 276.50 ± 24.58 0.62 ± 0.05 0.97 + 0.09 0.17 ± 0.02 0.24 ± 0.03 0.59 ± 0.04

30.00 £ 1.47 56.88 ± 2.90 390.60 + 22,73 0.53 ± 0.03 0.93 ± 0.05 0.29 ± 0.02 0.205:0.02 0.51 5:0.04

1. Orienti et al. / Journal of ControUed Release 31 (1994) 61-71 69

(A) (B)

oA

O

r,.

1,0-

0,8"

0 , 6

0 , 4

0,2 •

0,0 0 100 200 300 400 500

Time (h)

• 1,0

~ 0,8

~ 0,6

~ 0,4

o • ~ 0,2

r,. 0,0

100 200 300 400 500

Time (h)

Fig. 5. Fractional release of HCA from: (A) BSA microparticles, prepared in the presence of: ( • ) corn oil, (11) isopropyl alcohol, ( e ) n- butyl alcohol, in water at 37°C; and (B) BSA/CBP microparticles, prepared in the presence of: ( a ) corn oil, ( [ ] ) isopropyl alcohol, (©) n- butyl alcohol, in water at 37°C.

(A) (B) oooj £ ooooj 500 500 ~ . ~

450" 450 *O

300

150

0

'O

v

¢)

5 0 100 150 200 250 ~'~

Time (h)

300-

150"

50 100 150 200 250

Time (h)

Fig. 6. Permeability of HCA in: (A) BSA microparticles, prepared in the presence of: ( • ) corn oil, (11) isopropyl alcohol, (o) n-butyl alcohol, undergoing bulk erosion; and (B) B SA/CBP microparticles, prepared in the presence of: ( a ) corn oil, (I-1) isopropyl alcohol, (O) n-butyl alcohol, undergoing bulk erosion.

Eq. 11 therefore may account for the n values obtained in this period if dissolution controls the release. The kinetic analysis conducted in the third Cs decrease period up to M,/Moo=0.6, for those systems whose fractional release was even lower than 0.6, does not give linear correlations between In M,/M~ and In t. The increasing release rate observed in this period may therefore be attributed to the prevalence of the D increase over the Cs decrease, kD2 being > > kc3, sug- gesting a release controlled by the bulk erosion process.

Lasting and fractional release of the different release periods

Fig. 5 reports the release profiles from the different systems and shows that the fractional amount released at each time follows the sequence:

CO > CO/CBP > nBU > nBU/CBP > ISP> ISP/CBP

This behaviour can be correlated with the permeability of each system obtained by the relationship: P = DCs [ 18] and reported in Fig. 6. Fig. 5 and Table 3 show

70 1. Orienti et al. / Journal of Controlled Release 31 (1994) 61-71

that M-CO release 60% of their content during the first Fickian-Anomalous period and during part of the sec- ond period while M-nBu and M-ISP release 60% of their content during the first Fickian-Anomalous period, the second zero order period and during part of the third period. The duration of the Fickian-Anoma- lous period decreases from M-CO to M-nBu and M- ISP and is lower in the presence of CBP. The fractional amount released in this period is higher in M-CO than M-nBu and M-ISP and decreases in the presence of CBP. The duration of the following zero order period and the relative fractional amount released is higher for M-ISP than M-nBu and increases in the presence of CBP. This behaviour indicates that the systems with higher k~ values are characterized by a longer zero order period and shorter Fickian-Anomalous period. This can be attributed to a more rapid achievement of the suitable hydrolysis degree which triggers the con- formational arrangements able to slow down the hydro- philic trasformations of the degrading polymer and consequently to establish the zero order release. A rapid establishment of the zero order period and low initial permeability values (M-ISP) decreases the duration of the Fickian-Anomalous period and the relative amount released while increasing the duration of the zero order period and consequently allowing higher fractional releases during this period.

4. Conclusions

Loading and release from BSA microparticles pre- pared in the presence of different drug dispersing phases are strongly influenced by the ability of the dispersing phase to interact with the polymeric chain. The resulting modifications induced on the crosslinked polymeric structure influence the degradation pattern by a control on D increase and Cs decrease over time. The release is first Fickian-Anomalous, then approaches zero order and finally is exponentially cor- related to time. The duration of each kinetic period is correlated to the degradation constants characteristic of each system while the release rate is correlated to the permeability of the degrading system over time.

References

111 K.J. Widder, A.E. Senyei and D.F. Ranney, Magnetically responsive microspheres and other carriers for the biophysical targeting of antitumor agents, Adv. Pharmacol. Chemother., 16 (1979) 213-271.

[ 2 [ S. Fujimoto, F. Endoh, M. Miyazaki, R.D. Shrestha, K. Okui and Y. Morimoto, Intra-arterial administration of heated albumin microspheres containing mitomycin C to rabbits with VX-2 tumor, Jpn. J. Surg., 14 (3) (1984) 252-257.

[ 31 K.J. Widder and A.E. Senyei, Magnetic microspheres: a vehicle for selective targeting of drugs, Pharmac. Ther. 20 ( 1983 ) 377- 395.

141 L. Ilium and S.S. Davis, Targeting of drugs parenterally by use of microspheres, J. Parent. Sci. Technol. 36 (1982) 242-248.

151 S.P. Vyas, S. Bhatnagar, P.J. Gogoi and N.K. Jain, Preparation and characterization of HSA-propanolol microspheres for nasal administration, Int. J. Pharm., 69 ( 1991 ) 5-12.

16] PK.Gupta, C.T. Hung and D.G. Perrier. Albumin microspheres. I1. Effect of stabilization temperature on the release of adriamycin, Int. J. Pharm, 33 (1986) 147-153.

171 P.K. Gupta, C.T. Hung, F.C. Lam and D.G. Perrier, Albumin microspheres. II1. Synthesis and characterization of microspheres containing adriamycin and magnetile, Int. J. Pharm. 43 (1988) 167-177.

[81 Y. Nishioka, S. Kyotani, M. Okamura, Y. Mori, M. Miyazaki, K. Okazaki, S. Ohnishi, Y. Yamamoto and K. Ito, Preparation and evaluation of albumin microspheres and microcapsules containing cisplatin, Chem. Pharm Bull. 37, 5 (1989) 1399- 1400.

[91 J.J. Burger, E. Tomlinson, E.M.A. Mulder and J.G. McVie, Albumin microspheres for intra-arterial tumor targeting. 1. Pharmaceutical aspects, Int. J. Pharm. 23 (1985) 333 344.

[10] K Egbaria and M. Friedman, Sustained release albumin microspheres containing antibacterial drugs: effects of preparation conditions on kinetics of drug release, J. Controlled Release, 14 (1990) 79-94.

I 11 ] C. Jones, M.A. Burton and B.N. Gray, Albumin microspheres as vehicles for the sustained and controlled release of doxorubicin, J. Pharm. Pharmacol., 41 (1987) 813-816.

[ 121 K. Sugibayashi, Y. Morimoto, T. Nadai, Y. Kato, A. Hasegawa and T. Arita, Drug carrier properties of albumin microspheres entrapped 5-fluorouracil, Chem. Pharm. Bull., 27, 1 (1979) 204-209.

[131 1. Orienti and V. Zecchi, Progesterone loaded albumin microparticles, J. Controlled Release, 27 (1993) 1-7.

[14] R. Arshady, Albumin microspheres and microcapsules: methodology of manufacturing techniques, J. Controlled Release, 14 (1990) 111-113.

[15] B. Wichert and P. Rohdewald, A new method for the preparation of drug containing polylactic acid microparticles without using organic solvents, J. Controlled Release, 14 (1990) 269-283.

[ 161 A. Kishida, J.B. Dressman, S. Yoshioka, Y. Aso and Y. Takeda,

L Orienti et al./Journal of ControlledRelease 31 (1994) 61-71 71

Some determinants of morphology and release rate from poly(L)lactic acid microspheres, J. Controlled Release, 13 (1990) 83-90.

[17] J.M. Crank, The Mathematics of Diffusion. Claredon Press, Oxford (1975).

[18] J.T. Carstensen, Pharmaceutics of Solids and Solid Dosage Form, Wiley-Interscience, New York (1977).

[ 19] C. Chothia, Principles that determine the structure of proteins, Ann. Rev. Biochem., 53 (1984) 537-572.

[20] J.S. Richardson, The anatomy and taxonomy of protein structure, Adv. Protein Chem., 34 ( 1981 ) 167-339.

[21 ] E. Doelker, Cin6tique et m6canismes de la lib6ration control6e ~t partir des syst~mes polym6riques, in P. Bud, F. Puisieux, E. Doelker and J.P. Benoit (Eds.), Formes Pharmaceutiques Nouvelles, Lavoisier, Paris, 1985.

[22] J. Heller and R.W. Baker, Theory and practice of controlled drug delivery from bioerodible polymers, in Controlled Release of Bioactive Materials, R. Baker, Ed., Academic Press, New York, London, Sydney, San Francisco, 1980, p. 1.

[231 J. Heller, Zero order drug release from bioerodible polymers, in Recent Advances in Drug Delivery Systems, J.M. Anderson and S.W. Kim, Eds., Plenum Press, New York and London, 1984, p. 101.

[ 24] P.L. Ritger and N.A. Peppas, A simple equation for description of solute release. II. Fickian and anomalous release from swellable devices, J. Controlled Release, 5 (1987) 37-42.

[25] G.W. Sinclair and N.A. Peppas, Analysis of non-Fickian transport in polymers using simplified exponential expressions, J. Membrane Sci., 17 (1984) 329-331.

[26] R. Gurny and N.A. Peppas, Modelling of sustained release of water soluble drugs from porous, hydrophobic polymers, Biomaterials, 3 (1982) 27.

[27] S.K. Chandrasekaran and D.R. Paul, Dissolution-controlled transport from dispersed matrixes, J. Pharm. Sci. 71 (1982) 1399.

[28] H.B. Lee, J.D. Andrade and M.S. Jhon, Nature of water in synthetic gels. II. Proton pulse NMR of polyhydroxy-ethyl methacrylate, Polym. Prepr. 13 (1972) 706.