In vitro and in vivo evaluation of polyoxyethylene indomethacin esters as dermal prodrugs

journal of ELSEVIER

Journal of Controlled Release 34 ( 1995 ) 223-232

controlled release

In vitro and in vivo evaluation of polyoxyethylene indomethacin esters as dermal prodrugs F.P. Bonina a,., L. Montenegro a, p. De Caprariis b, F. Palagiano b, G. Trapani c, G. Liso c a Institute of Pharmaceutical and Toxicological Chemistry, Viale A Doria 6, Universitgt di Catania, 95125 Catania, Italy b Department of Pharmaceutical Chemistry, University ofNapoli, Napoli, Italy c Department of Pharmaceutical Chemistry, University ofBari, Bari, Italy Received 9 June 1994; accepted 22 December 1994

Abstract Indomethacin polyoxyethylene esters (1-5) were synthesized and assessed both in vitro and in vivo for their usefulness as indomethacin dermal prodrugs. Esters 1-5 showed good water stability and rapid enzymatic cleavage and their hydrolysis rates, both chemical and enzymatic, were not significantly affected by the lengthen of the polyoxyethylenic chain used as promoiety. Results from in vitro percutaneous absorption studies showed that only esters 4 and 5 significantly increased indomethacin cumulative amount penetrated through excised human skin compared to the parent drug. In vivo topical anti-inflammatory activity, using methyl nicotinate (MN) induced erythema as inflammation model in human volunteers, was investigated for ester 5, which showed the best in vitro results. In vivo results showed an interesting delayed and sustained activity of ester 5 compared to the parent drug. Keywords: Indomethacin; Polyoxyethylene glycol; Dermal prodrug; In vivo; Sustained activity

1. Introduction The therapeutic efficacy of a drug, following its application onto the skin, mainly depends on its ability to penetrate the skin at such extent to elicit the desired pharmacologic activity. Since most drugs show unsuitable physicochemical properties to penetrate effectively the skin, different strategies have been developed to increase drug skin permeation. Among these strategies, in addition to the use of penetration enhancers, the prodrug approach is one of the most promising. For a successful dermal prodrug approach [ 1 ], the prodrug should exhibit the following features: (1) adequate * Corresponding author. 0168-3659/95/$09.50 © 1995 Elsevier Science B.V. All rights reserved SSDIO1 6 8 - 3 6 5 9 ( 9 5 ) 0 0 0 0 3 -8

aqueous stability such that its formulation in practical vehicle is possible; (2) controlled enzymatic conversion into the parent drug within the viable tissue; (3) enhanced biphasic (both lipophilic and aqueous) solubility [2,3]. Indomethacin is a potent lipophilic antiinflammatory drug whose topical therapeutic use is strongly limited by its inability to penetrate the skin [4]. Many papers report the use of skin penetration enhancers to increase indomethacin skin permeation [ 5 - 7 ] but only few reports have been published [8,9] on the use of the prodrug approach with this aim. Recently, we synthesized indomethacin dermal prodrugs using N-alkyllactams, which possess inherent enhancing ability, as promoieties [ 10]. Indomethacin

224

F.P. Bonina et al. / Journal of Controlled Release 34 (1995) 223-232

N-alkyllactam esters showed the main requirements, such as good water stability, rapid enzymatic hydrolysis and increased flux through excised human skin, to be regarded as interesting indomethacin dermal prodrugs. In this work, we synthesized the indomethacin polyoxyethylene esters 1-5 (see Fig. 1) in order to evaluate the possibility of employing different new promoieties in designing indomethacin dermal prodrugs. Polyoxyethylene glycols were chosen as promoieties since: (a) they are common constituents of topical drug formulations; (b) they possess inherent skin penetration enhancing ability; (c) they are hydrophilic and should impart to the prodrug a suitable water solubility, in addition to an increased lipophilicity due to the esterification of indomethacin carboxylic group. So, in this paper we assessed the chemical and enzymatic hydrolysis and the flux through excised human skin of indomethacin polyoxyethylene esters 1-5 to evaluate their usefulness as indomethacin dermal prodrugs. Furthermore, since the results obtained from in vitro experiments were encouraging, in order to investigate the correlation between in vitro skin permeation enhancement and in vivo topical anti-inflammatory activity, the latter was investigated for the derivative 5 which showed the best skin penetration ability. Methyl nicotinate (MN) induced erythema, on healthy human volunteers, was chosen as inflammation model and the extent of erythema was monitored by the non-invasive technique of reflectance spectrophotometry. The in vivo topical anti-inflammatory activity of ester 5 and indomethacin are reported and discussed in this paper.

2. Materials and methods

CH2-CO0-(CH2-CH2-O) n-CH2-CH2-OH CH30~J~/~ I

cl~ ~ c F i g . 1. C h e m i c a l

~ o

s t r u c t u r e o f e s t e r s 1 - 5 . E s t e r 1 : n = 1 ; e s t e r 2: n = 2 :

e s t e r 3: n = 3; e s t e r 4 : n = 4 ; e s t e r 5: n = 5.

of a Waters 600 pump with a model 486 UV-Vis detector, an automatic sample injection module Wisp model 712, a Waters C 18/xBondapack 4.6 mm × 30 cm reverse phase column and a Nec PowerMate SX Plus computer. Indomethacin was purchased from Sigma (St. Louis, MO). Diethylene glycol, Triethylene glycol, Tetraethylene glycol, Pentaethylene glycol, Hexaethylene glycol and Methyl nicotinate (MN) were obtained from Fluka (Switzerland). Acetonitrile and water used in the HPLC procedures were of LC grade and were obtained from Carlo Erba (Italy). All other chemicals were of reagent grade.

2.2. Synthesis of indomethacin esters 1-5 Esters 1-5 were synthesized using the method reported by Cecchi et al. [11] for synthesizing poly(oxyethylene) esters of ibuprofen. The products were purified by flash chromatography using chloroform as eluant for compound 1 and ethyl acetate for compounds 2-5. The products obtained were all oils and they failed to crystallize, except the compound 1, which was crystallized from n-hexane (mp 92-93°C). Elemental analyses (C,H,N) were within ___0.3% of the theoretical values. The I.R. and 'H-NMR spectral data of esters 1-5 were consistent with their chemical structures.

2.1. Materials 2.3. Chemical and enzymatic hydrolysis Melting points were taken on Buchi 510 capillary melting point apparatus and were uncorrected. The IR spectra were recorded on a Perkin Elmer infrared spectrophotometer model 281 using sodium chloride plates for neat liquid compounds and potassium bromide plates for the solid compound. 1H-NMR spectra were recorded on a Bruker WM 250 using CDC13 as solvent. Elemental analysis was performed on a Carlo Erba model 1108 elemental analyzer. The HPLC consisted

The chemical hydrolysis rate of ester 1-5 was determined in solution of isotonic phosphate buffer pH 7.4, at 32°C, following the disappearance of the ester by the HPLC method described below. Enzymatic hydrolysis rate of esters 1-5 was determined in human plasma. Plasma fractions (4 ml) were diluted with 1 ml of isotonic phosphate buffer pH 7.4 (80% plasma). The reaction was started by adding 100/xl of a stock solu-

F.P. Bonina et al. / Journal of Controlled Release34 (1995) 223-232 tion of esters 1-5 in methanol ( 1.0 mg/ml) to 5 ml of plasma prethermostatted at 37 + 0.2°C. Plasma samples were thermostatted at 37 + 0.2°C during the experiments. Aliquots (300 ~1) were withdrawn at intervals and deproteinized by adding 600/~1 of 0.01 M HC1 in methanol. After centrifugation at 5000 × g for 5 min, 25 ~1 of the clear supernatant was analyzed for indomethacin or ester 1-5 concentration, using the HPLC method described below. Pseudo-first order rate constants for the chemical and enzymatic hydrolysis were calculated from the slopes of linear plots of the logarithm of residual indomethacin ester against time. 2.4. Partition coefficient o f indomethacin and esters 1-5

The partition coefficients n-octanol/water of indomethacin and its esters 1-5 were determined at ambient temperature, according to the method described by Inagi et al. [ 12]. The solute concentrations in the aqueous phase, before and after partition, were determined by the HPLC method reported below. 2.5. In vitro skin permeability Samples of adult human skin (mean age 43 + 8 years) were obtained from breast reduction operations. Subcutaneous fat was removed and the skin was immersed in distilled water at 60 + 1°C for 2 min. [ 13 ], then stratum corneum and epidermis (SCE) were peeled off. SCE membranes were dried in a desiccator at approx. 25% RH and then wrapped in aluminium foil and stored at 4 + I°C until use [ 14]. Dried SCE samples were rehydrated by immersion in distilled water, at room temperature, for 1 h before being mounted in Franz-type diffusion cells ( LGA, Berkeley, CA). The skin surface available for absorption was 0.75 cm 2 and the receptor volume was 4.5 ml. The receiving compartment contained ethanol/water 50:50 to ensure sink conditions [ 15,16]. The receiving solution was stirred and maintained at 35 + 1°C throughout the experiments. Indomethacin and its esters 1-5 were dissolved in ethanol (5 mg/ml) and 200/xl was placed on the skin surface. The solvent was allowed to evaporate and the experiment was run for 24 h. Samples of the receiving solution (50/xl) were withdrawn at 24 h and analyzed for indomethacin or esters 1-5 content by the HPLC method described below.

225

2.6. H P L C analysis o f indomethacin and esters 1-5 Indomethacin and esters 1-5 were determined by HPLC using a convex gradient starting with acetonitrile/acetic acid 0.1 M 40:60, changing to acetonitrile/ 0.1 M sodium acetate 40:60 over 5 min, then changing to acetonitrile/0.1 M sodium acetate 60:40 over 10 min and then returning to the initial conditions over 10 min. The flow rate was 1.8 ml/min and the effluent was continuously monitored at 250 nm. The retention times were: indomethacin, 8.5 rain; ester 1, 11.9 min; ester 2, 12.2 min; ester 3, 12.4 min; ester 4, 12.6 min; ester 5, 12.7 min 2. 7. Preparation of aqueous gel Carbopol gels, containing indomethacin or ester 5, were prepared by dispersing Carbopol 934 ( 1.5% w/ w) in distilled water (75.4% w / w for indomethacin gel and 74.7% w / w for ester 5 gel) with constant stirring. Indomethacin ( 1.0% w/w, 0.0027 mol) or ester 5 ( 1.7% w/w, 0.0027 mol) were solubilized in ethanol (20.0% w/w) together with methyl-p-hydroxybenzoate (0.1% w/w). The ethanolic solution was added to the carbopol dispersion and the mixture was then neutralized and made viscous by the addition of triethanolamine (2.0% w/w). The gels were stored at room temperature for 24 h under air-tight conditions prior to use. 2.8. In vivo anti-inflammatory activity o f ester 5 on methyl nicotinate induced erythema Instrument The induced MN erythema was monitored by using a reflectance visible spectrophotometer X-Rite model 968, having 0 ° illumination and 45 ° viewing angle. The instrument was calibrated with a supplied white standard traceable to the National Bureau of Standards' perfect white diffuser. The spectrophotometer was controlled by an IBM PS2 50 computer, which performed all color calculations from the spectral data by means of a menu-driven suite of programs (Spectrostart) supplied with the instrument. Reflectance spectra were obtained over the wavelength range 400-700 nm using illuminant C and 2° standard observer. From the spectral data obtained, the erythema index (E.I.) was cal-

226

F.P. Bonina et al. /Journal of Controlled Release 34 (1995) 223-232

Table 1 Molecularweight (MW), chemicaland enzymatichydrolysis (t~/2). partitioncoefficient( log P I, calculatedwater solubility (Sw) and cumulative amount penetrating through excised human skin after 24 h (Q) of indomethacin and esters 1-5 Compound

Indomethacin 1 2 3 4 5

MW

357.8 445.9 490.0 534.0 578.1 622.2

t,/2 (ht

Buffer pH 7.4

Humanplasma

533 695 540 621 562

3.40 3.82 3.15 3.56 2.81

Log P (n-octanol/water)

Sw (/xg/mll

Q_+ S.D? (p~g/cmz)

3.10 4.52 4.39 4.15 3.99 3.81

82 18 45 89 143 240

3.00 + 0.37 1.80 + 0.66 3.20 + 0.78 3.89_+0.59 16.31 +2.01 19.42_+ 1.89

aEach experiment was run in duplicate on three different donors. culated using an equation (Eq. 1 ) similar to that reported by Dawson et al. [ 17]: E . I . = 1 0 0 ~ L o g ~ l + 1.5(LOgRs~o+ Log --~-1] L 8560 R58o] - 2(Log ~ 1 + L o g ~ l ] ] \ Rslo R61o]J

(1)

where ( l / R ) is the inverse reflectance at a specific wavelength (560, 540, 580, 510, 610 nm). Protocol

In vivo experiments were performed on six volunteers of both sexes in the age range 25-35 years. The volunteer subjects were fully informed of the nature of the study and the procedures involved. The participants did not suffer from any ailment and were not on any medication at the time of the study. They were rested for 15 min prior to the experiments and room conditions were set at 22 + 2C ° and 4 0 - 5 0 % relative humidity. Eight sites on the ventral surface of each forearm were defined using a circular template (1 cm 2) and demarcated with permanent ink. For each volunteer, two of the eight sites of each forearm were used as controls applying 50 mg of gel without active compounds and the other six sites were treated with 50 mg of indomethacin or ester 5 gel. The preparations and were spread uniformly on the site by means of a solid glass rod. The sites were then occluded for 3 h, using Hill Top chambers (Hill Top Research, Inc., Cincinnati, O H ) . After the occlusion period, the chambers were removed and the skin surfaces were washed to remove the gel and allowed to dry for 15 min On each pretreated site, M N aqueous solution (0.5% w / v ) was

applied at different times after gel removal: immediately (t = 0), 3 and 6 h later (t = 3 and t = 6, respectively). MN was applied on the skin surface for 1 min, using a Hill Top chamber ( 1 cm z) whose cotton pad was saturated with 200 /xl of MN solution and the induced erythema was monitored for 100 min. E.I. baseline values were taken at each designated site before application of gel formulations and they were subtracted from the E.I values obtained after MN application at each time point, to obtain AE.I. values. For each site, the area under the response ( A E . I . ) - t i m e curve [ AUC ] was computed using the trapezoidal rule.

3. Results and discussion 3.1. C h e m i c a l a n d e n z y m a t i c h y d r o l y s i s

As shown in Table 1, esters 1-5 had a notable stability in phosphate buffer at pH 7.4 and no significant difference in their hydrolysis rates was observed as the length of the polyoxyethylene chain increase. An essential prerequisite for success in the use of prodrugs is that prodrug reconversion into the parent drug occurs in the skin. Since the preparation of skin homogenates may present some problems due to the tenacious and elastic nature of the outermost layer of the skin [ 18] different models have been developed to mimic skin esterase activity and to assess the susceptibility of ester prodrugs in undergoing bioconversion in the skin. So, many authors [4,18,19] reported the possibility of using human plasma to assess the hydrolysis rates of ester prodrugs for dermal delivery.

F.P. Bonina et al. / Journal of Controlled Release 34 (1995) 223-232

227

5.0

4.0

i

i

N u m b e r of p o l y o x y e t h y l e n i c

groups

Fig. 2. Log P vs. numberof polyoxyethylenicgroups of esters 1-5. As may be noted in Table l, all the esters were hydrolyzed by human plasma and no significant difference of ester 1-5 half-lives were observed. From these results a rapid enzymatic hydrolysis of esters 1-5 within the viable skin could not be expected. However, some authors found a significant greater enzymatic hydrolysis within the skin that in human plasma for nalidixic acid dermal prodrugs [19]. Furthermore, other authors [4] reported successful dermal prodrugs with enzymatic hydrolysis rates similar or slower than those of esters 1-5.

3.2. Lipophilicity and water solubility Stratum corneum is generally believed the main bartier in drug skin permeation process. Since the horny layer is basically a lipophilic barrier, drug lipophilicity is regarded as one of the key parameters, but not the only, which controls drug skin permeation. So, one of the main objectives in dermal prodrug design is to obtain prodrug with increased lipophilicity compared to the parent drug. Many authors [2,20] have outlined that more lipophilic drug derivatives could show better partitioning and solubility into the SC which could result in enhanced skin permeation. Regarding indomethacin esters synthesized in this work, esterification of indomethacin carboxylic group gave, as expected, esters with increased lipophilicity compared to the parent drug (see Table 1 ).

The indomethacin Log P value, reported in Table 1, agreed well with that reported by other authors [ 4]. As may be noted in Fig. 2, ester I-5 lipophilicity decreased as the polyoxyethylene chain lengthened and this pattern is in agreement with the increasing hydrophilicity introduced in these molecules by increasing the number of polyoxyethylenic groups. Many authors [ 1,2], on the basis of theoretical considerations and experimental data, have pointed out that dermal prodrug water solubility may play a role as important as lipophilicity in skin permeation process, especially for very lipophilic drugs. Since indomethacin esters 1-5 were all oils with the exception of ester 1 it is difficult to measure experimentally their water solubility. So, we calculated this parameter for esters 1-5 by using the theoretical method reported by Yalkowsky et al. [ 21,22 ]. According to this method, which has already been used in dermal prodrug design and evaluation [ 23 ], two different equations are used to calculate water solubility of solid compounds (Eq. 2) and oils (Eq. 3). Log Sw = - l o g P - 0 . 0 1 M P + 1.05

(2)

Log Sw = - 1.072 log P + 0.672

(3)

where Sw represents compound water solubility and MP is its melting point. Indomethacin and ester 1-5 water solubility, calculated using Eq. 2 and 3, are reported in Table 1. Notwithstanding that indomethacin water solubility calculated using Eq. 2 did not agree well with that previously and experimentally determined by us [ 10], we believe that water solubility values calculated by means of these equations could be useful to compare indomethacin and ester 1-5 water solubility and to evaluate the effect of this parameter on their in vitro skin permeation. As may be noted in Table 1, ester 1-3 water solubility was lower (esters 1,2) or similar (ester 3) to that of indomethacin while a linear increase of this parameter was observed for esters 4 and 5 compared to the parent drug. This pattern is in agreement with the increasing hydrophilic character introduced in these molecules by lengthening the polyoxyethylenic chain.

3.3. In vitro skin permeability In vitro skin permeation results, expressed as indomethacin or indomethacin equivalent cumulative amount penetrated through human skin after 24 h, are reported in Table 1. As may be noted, esters 1-3 did

228

F.P. Bonina et al. / Journal of Controlled Release 34 (1995) 223-232 50

The data obtained studying in vitro percutaneous absorption of esters 1-5, using SCE membranes, confirm the importance of increasing biphasic solubility (both in water and in oil) in dermal prodrug design. At the same time, these data agree well with the theory, reported by Sloan [20] and accepted by other authors [2], that in a homologous series of prodrugs with

c O t/

_u

I)

.>

25

(A) 20

Foo

~;o

~oo' ~;o

o6o

i

~o

~oo

15

W a v e l e n g t h (nrn)

Fig. 3. Skin reflectance spectra. Curve a: normal skin: curve b: MN induced skin erythema.

not significantly increase indomethacin cumulative amount penetrated through the skin while applying esters 4 and 5 onto the skin the cumulative amount of indomethacin penetrated was 5.5- and 6.5-fold higher, respectively, compared to that obtained after topical application of indomethacin. In these experiments no hydrolysis of the esters during the skin permeation process was observed. Probably, the lack of enzymatic hydrolysis that we found in our in vitro permeation studies is due to the use of SCE membranes obtained by means of a thermal separation technique and to the large amounts of ethanol used in the receptor compartment. Carrying out some preliminary in vitro skin permeation experiments (unpublished data) using full thickness human skin (fresh excised) we obtained much lower fluxes (almost undetectable) applying both indomethacin or esters 1-5 on the skin. However, in this last experiments, only indomethacin was detected in the receptor phase, thus indicating a complete hydrolysis of the esters within 24 h. Other authors [24] reported that hydrolytic enzymes are rapidly leached out from whole human skin samples during the permeation study and this may lead to pronounced drug metabolism in the receptor phase. The lower fluxes obtained using whole skin may be explained on the basis of the evidence reported by others [25,26] that the dermis in vitro can act as a significant additional artificial barrier to the absorption of lipophilic compounds.

LLJ <3

10

~'o

~

6'o

~o

16o

120

rain 25

(B) 2C

15

"-.... \

,4 W

<:~ 10

5

C 0

2b

1o

~o

1o

6o

120

rain

Fig. 4. (a) AE.I. values vs. time for one subject. Indomethacin or ester 5 gels were applied for 3 h and MN was applied (a) immediately after their removal ( t = 0 ) ; or (b) 6 h after their removal ( t = 0 ) . ( v ) Indomethacin; ( O ) Ester 5, ( • ) Control.

F.P. Bonina et al. / Journal of Controlled Release 34 (1995) 223-232

229

Table 2 AUCo_mo values obtained pretreating skin sites with gel containing indomethacin or ester 5 and applying MN immediately after gel removal ( t = 0 ) , 3 h ( t = 3 ) o r 6 h ( t = 6 ) later Subject

AUC~_loo

Control

t=0

A B C D E F Mean +S.D.

t=3

t=6

Indomethacin

Ester 5

Indomethacin

Ester 5

Indomethacin

Ester 5

395.4 526.8 341.1 402.6 509.4 467.0 440.4 +72.4

817.3 1011.9 604.3 756.1 988.7 874.2 842.5 + 152.2

588.7 793.0 553.5 664.7 764.2 503.1 644.5 + 116.7

579.3 857.6 591.7 702.5 461.4 498.5 615.2 + 145.2

829.8 1025.2 586.7 1149.3 821.7 954.4 894.5 ± 194.8

529.3 761.5 327.0 736.1 501.7 467.9 553.9 + 166.4

1083.4 1231.1 808.0 1093.9 894.5 1174.5 1047.6 + 163.7

aEach value represents the mean of two different sites in the same subject.

increased lipophilicity compared to the parent drug the highest flux through the skin is achieved by the derivative which shows the highest water solubility. 3.4. In vivo anti-inflammatory activity of indomethacin and ester 5 Since the results obtained in vitro were very promising, we thought it noteworthy to investigate the topical anti-inflammatory activity of derivative 5 which had shown the best in vitro results. So indomethacin and ester 5 carbopol gel were prepared and used for evaluating their topical anti-inflammatory activity. Different models have been reported in the literature for evaluating topical anti-inflammatory activity of NSAIDs [ 27 ]. Among these models, UV-B and methyl nicotinate induced erythema are the most used in humans [28,29]. For decades, the most common method for evaluating skin erythema has been visual assessment by trained observers. This method is affected by large variations of interindividual and interobserver evaluations. Recently, non-invasive techniques such as Laser Doppler Flowmetry, reflectance colorimetry or spectrophotometry have been used to obtain objective and more accurate evaluations of skin erythema [ 30-32]. Reflectance filter colorimetry has been extensively used [ 33,34 ] for designating the extent of erythema by measuring the skin color surface in terms of CIE (Commission International d'Eclairage) L* a* b* color space

parameters since some authors [35,36] found significantly correlation between a* values and visual grading of skin erythema. Unlikely reflectance colorimetry, reflectance spectrophotometry provides skin reflectance spectra, generally in the range 400-700 nm, from which it is possible to obtain erythema index (E.I.) values (see Eq. 1 ) for more accurate and reliable evaluations of skin erythema [ 17,37,38]. The E.I. values were calculated, at each site and at different times, to monitor the extent of MN induced skin erythema. In Fig. 3, typical skin spectra before (curve a) and after (curve b) MN application on the skin show the notable increase of hemoglobin absorption peaks mainly between 510 and 590 nm. Since the bioavailability of MN in solution is about 10 to 100 times more rapid than that of NSAIDs such as indomethacin [39], we pretreated the skin with gel containing Indomethacin or ester 5, before MN application. A 3-h pretreatment time, under occlusion, was chosen since Nowack et al. [39], studying the percutaneous bioavailability of 1% indomethacin gel applied under occlusion in human, reported that drug maximal concentration was obtained at between 2 and 4 h. A typical time course of MN induced erythema on skin sites pretreated with indomethacin or ester 5 gels, for one subject, is reported in Fig. 4. Plotting AE.I. values versus time, AUC values were determined, for each subject, by calculating the areas between the response curve and the x-axis, and the mean AUC values are reported in Table 2. AUC values were inversely related

230

F.P. Bonina et al. / Journal of Controlled Release 34 (1995) 223-232

60

*/. inhibition /

50 40 30

g

10 0

t=O

r/i [ t=3

[~-ltndomethacin m

t:6

Ester 5

Fig. 5. Indomethacin and ester 5 percentage inhibitions of the inflammatory reaction induced by MN.

to indomethacin and ester 5 ability to inhibit MN induced skin erythema. As may be noted in Table 2, a time-dependent profile of activities was observed applying MN at different time ( t = 0 h, t = 3 , t = 6 h) after active compound removal. At t = 0 indomethacin was more effective than ester 5 in inhibiting the induced erythema. When MN was applied 3 h after having removed the gels, indomethacin efficacy was not significantly different from that observed for ester 5 and at t = 6 h ester 5 showed a significantly ( p < 0 . 0 5 ) greater inhibitory activity than indomethacin. As reported by other authors [40] in vivo topical anti-inflammatory activity evaluation of NSAIDs using a noninvasive technique, to better outline the results obtained it is possible from [ AUC ] values to calculate the percentage inhibition of the inflammatory reaction induced by MN: Inhibition (%) _ AUC(c) -- A U C ( T ) AUC~c)

X 100

low ( about 19% ) and increased at t = 3 h ( about 40%), remaining constant when MN was applied 6 h after gel removal. This delayed and sustained topical antiinflammatory activity of ester 5 could be explained on the basis of a greater SC reservoir, due to its increased lipophilicity compared to the parent drug, and a slow release from SC to underlying aqueous tissue modulated by its enzymatic hydrolysis rate. The greater in vitro skin permeation of ester 5 with respect to the parent drug does not agree well with the anti-inflammatory topical results obtained inducing erythema immediately after gel removal (t = 0) while they are in accordance with those obtained at t = 6. The lack of correlation between in vitro permeability and in vivo efficacy at t = 0 could be likely due to a slow prodrug hydrolysis within the skin. The anti-inflammatory results obtained in this work suggest that it could be important to evaluate in vivo dermal prodrug efficacy at different times after their skin application since a suitable time could be needed for regenerating and achieving skin therapeutic concentrations of the parent drug. In conclusion, polyoxyethylene glycols resulted interesting promoieties for indomethacin dermal prodrug design since esters 1-5 showed the main requirements needed for dermal prodrugs such as chemical stability, enzymatic lability and increased in vitro skin permeation. The evaluation of ester 5 in vivo topical anti-inflammatory activity, using MN induced erythema in human volunteers as inflammation model, has pointed out an interesting delayed and sustained activity compared to the parent drug.

(4)

where AUC
Acknowledgements We thank Assessorato Beni Culturali, Regione Sicilia for financial support.

References [1] K.B. Sloan (Ed.), Prodrugs: Topical and Ocular Drug Delivery, Marcel Dekker, New York, 1992. [2] R.H. Guy and J. Hadgraft, Percutaneous penetration enhancement: Physicochemical considerations and implications for prodrug design, in: K.B. Sloan (Ed.), Prodrugs: Topical and Ocular Drug Delivery, Marcel Dekker, New York, 1992, pp. 1-16.

F.P. Bonina et al. / Journal of Controlled Release 34 (1995) 223-232

[ 3 ] D. Friend, P. Catz, J. Heller, J. Reid and R. Baker, Transdermal delivery of levonorgestrel II: effect of prodrug structure on skin permeability in vitro, J. Control. Release 7 (1988) 251-261. [4] G.B. Kasting, R.L. Smith, and B.D. Anderson, Prodrugs for dermal delivery: solubility, molecular size, and functional group effects, in: K.B. Sloan (Ed.), Prodrugs: Topical and Ocular Drug Delivery, Marcel Dekker, New York, 1992, pp. 117-161. [5] T. Ogiso, Y. Ito, M. Iwaki and H. Atago, Absorption of indomethacin and its calcium salt through rat skin: effect of penetration enhancers and relationship between in vivo and in vitro penetration, J. Pharmacobio-Dyn. 9 (1986) 517-525. I6] H. Okabe. K. Takayama, A. Ogura and T. Nagai, Effect of limonene and related compounds on the percutaneous absorption of indomethacin, Drug Des. Del. 4 (1989) 313321. [7] J. Catz and D.R. Friend, Alkyl esters as skin permeation enhancers for indomethacin, Int. J. Pharm. 55 (1989) 17-23. [8] K.B. Sloan, S. Selk, J. Haslam, L. Caldwell and R. Shaffer, Acyloxyamines as prodrugs of anti-inflammatory carboxylic acids for improved delivery through the skin, J. Pharm. Sci. 73 (1984) 1734-1737. [9] S.M. Milosovich, A.A. Hussain, M. Hussain and L. Dittert, The utilization of prodrugs to enhance transdermal absorption of testosterone, deoxycorticosterone and indomethacin, Prog. Clin. Biol. Res. 292 (1989) 272-277. [ 10] F.P. Bonina, L. Montenegro, P. De Caprariis, E. Bousquet, S. Tirendi, l-Alkylazacycloalkan-2-one esters as prodrugs of indomethacin for improved delivery through human skin, Int. J. Pharm. 77 ( 1991 ) 21-29. [ 11 ] R. Cecchi, L. Rusconi, M.C. Tanzi, F. Danusso and P. Ferruti, Synthesis and pharmacological evaluation of poly(oxyethylene) derivatives of 4-isobutylphenyl-2propionic acid (ibuprofen), J. Med. Chem. 24 (1981) 622625. [ 12] T. Inagi, T. Muramatsu, H. Nagai and H. Terada, Mechanism of indomethacin partition between n-octanol and water, Chem. Pharm. Bull. 29 ( 1981 ) 2330-2337. [ 13] A.M. Kligman and E. Christophers, Preparation of isolated sheets of human skin, Arch. Dermatol. 88 (1963) 702-705. [14] J. Swarbrick, G. Lee, J. Brom and N.P. Gensmantel, Drug permeation through human skin I: effect of storage conditions of skin, J. Invest. Dermatol. 78 (1982) 63-66. [ 151 L.G. Mueller, Novel anti-inflammatory esters, pharmaceutical composition and methods for reducing inflammation, UK Patent, GB 2 204 869 A, 23 Nov. 1988. [ 16] E. Touitou and B. Fabin. Altered skin permeation of a highly lipophilic molecule: tetrahydrocannabinol, Int. J. Pharm. 43 (1988) 17-22. [ 17] J.B. Dawson, D.J. Barker, D.J. Ellis, E. Grassam, J.A. Catterill, G.W. Fisher and J.W. Feather, A theoretical and experimental study of light absorption and scattering by 'in vivo' skin, Phys. Med. Biol. 25 (1980) 696-709. [ 18] M. Johansen, B. Mollgaard, P.K. Wotton, C. Larsen and A. Hoelgaard, In vitro evaluation of dermal prodrug deliverytransport and bioconversion of a series of aliphatic esters of metronidazole, Int. J. Pharm. 32 (1986) 199-206.

231

[ 19 ] H. Bundgaard, N. Mork and A. Hoelgaard, Enhanced delivery of nalidixic acid through human skin via acyloxymethyl ester prodrugs, Int. J. Pharm. 55 (1989) 91-97. [ 20] K.B. Sloan, Prodrugs for dermal delivery, Adv. Drug Del. Rev. 3 (1989) 67-101. [ 21 ] S.H. Yalkowsky and S.C. Valvani, Solubility and partitioning I: solubility of uonelectrolytes in water, J. Pharm. Sci. 69 (1980) 912-922. [22] S.H. Yalkowsky, S.C. Valvani and T.J. Roseman, Solubility and partitioning VI: octanol solubility and octanol-water partition coefficients, J. Pharm. Sci. 72 (1983) 866-870. [23] D.W. Osborne and W.J. Lambert, A computational method for predicting optimization of prodrugs or analogues designed for percutaneous delivery, in: K.B. Sloan (Ed.), Prodrugs: Topical and Ocular Drug Delivery, Marcel Dekker, New York, 1992, pp. 163-177. [24] H. Bundgaard, Hoelgaard A. and B. Mollgaard. Leaching of hydrolytic enzymes from human skin in cutaneous permeation studies as determined with metronidazole and 5-fluorouracil pro-drugs, Int. J. Pharm. 15 (1983) 285-292. [25] R.J. Scheuplein and I.H. Blank, Mechanism of percutaneous absorption, IV. Penetration of non-electrolytes (alcohols) from aqueous solution and from pure liquid, J. Invest. Dermatol. 60 (1973) 286-296. [26] R.L. Bronaugh and R.F. Stewart, Methods for in vitro percutaneous absorption studies. III. Hydrophobic compounds, J. Pharm. Sci. 73 (1984) 1255-1258. [27] M. Bouclier, A. Chatelus and C.N. Hensby, In vivo animal models for the evaluation of anti-inflammatory drug action in the skin, in: N.J. Lowe and C.N. Hensby (Eds.), Nonsteroidal Anti-inflammatory Drugs, Karger, Basel, 1989, pp. 118-132. [28] P.M. Farr and B.L. Diffey, A quantitative study of the effect of topical indomethacin on cutaneous erythema induced by UVB and UVC radiation, Br. J. Dermatol. 115 (1986) 453-466. [29] S.Y. Chan and A. Li Wan Po, Quantitative assessment of nonsteroidal anti-inflammatory topical products in nicotinateinduced erythema using tristimulus colour analysis, Int. J. Pharm. 83 (1992) 73-86. [301 P.H. Andersen, K. Abrams, P. Bjerring and H.A. Maibach, A time-correlation study of ultraviolet B-induced erythema measured by reflectance spectroscopy and laser Doppler flowmetry, Photodermatol. Photoimmunol. Photomed. 8 (1991) 123-128. [311 J.W. Feather M. Hajizadeh-Saffar, G. Leslie and J.B. Dawson, A portable scanning reflectance spectrophotometer using visible wavelenghts for the rapid measurements of skin pigments, Phys. Med. Biol. 34 (1989) 807-820. [32] J.C. Seitz and C.G. Whitmore, Measurements of erythema and tanning responses in human skin using a tristimulus colorimeter, Dermatologica 177 (1988) 70-75. [331 W. WesterhoL B.A.A.M. van Hasselt and A. Kammeijer, Quantification of UV-induced erythema with a portable computer controlled chromameter, Photodermatology 3 (1986) 310-314. [34] I.M. Gibson, Measurements of skin color in vivo, J. Soc. Cosmet. Chem. 22 (1971) 725-740.

232

F.P. Bonina et al. /Journal of Controlled Release 34 (1995) 223-232

[35] N. Muizzudin, K. Marenus, D. Maes and W.P Smith, Use of chromameter in assessing the efficacy of anti-irritants and tanning accelerators, J. Soc. Cosm. Chem. 41 (1990) 369-378. [ 36 ] E.H. Braue Jr., M.M. Mershon, J.V. Wade and M.R. Litchfield, In vivo assessment of vescicant skin injury using Minolta Chroma Meter, J. Soc. Cosmet. Chem. 41 (1990) 259-265. [37] R.R. Anderson and J.A. Parrish, The optics of human skin, J Invest. Dermatol. 77 ( 1981 ) 13-19.

[ 38] P.H. Andersen and P. Bjerring, Spectral reflectance of human skin in vivo, Photodermatol. Photoimmunol. Photomed. 7 (1990) 5-12. [39] H. Nowack, U. Matin, R. Reger, H. Bohme, K.H. Schriever, P. Bocionek, R. Elbers and H.G. Hampfmeyer, Cutaneous absorption of indomethacin from two topical preparations in volunteers, Pharm. Res. 2 (1985) 202-206. [ 40] M.C. Poelman, B. Piot, F. Guyon, M. Deroni and J.L. Leveque, Assessment of topical non-steroidal anti-inflammatory drugs, J. Pharm. Pharmacol. 41 (1989) 720-722.