Iontophoretic delivery of 5-aminolevulinic acid and its methyl ester using a carbopol gel as vehicle

Journal of Controlled Release 98 (2004) 57 – 65 www.elsevier.com/locate/jconrel

Iontophoretic delivery of 5-aminolevulinic acid and its methyl ester using a carbopol gel as vehicle Nadia Merclin *, Tobias Bramer, Katarina Edsman Department of Pharmacy, Uppsala Biomedical Centre, Uppsala University, P.O. Box 580, Uppsala SE-751 23, Sweden Received 28 November 2003; accepted 22 April 2004 Available online 8 June 2004

Abstract The aim of this study was to evaluate a Carbopol gel as a vehicle for iontophoretic delivery of 5-aminolevulinic acid (ALA) and its methyl ester (m-ALA). The formulation was characterized rheologically and the passive diffusion of ALA and m-ALA in the gels was measured. Addition of ALA and m-ALA did not change the rheological behavior of the gel and the diffusion coefficients of ALA and m-ALA were 4.4 F 1.2
10 6 and 3.08 F 0.7
10 7 cm2 s 1, respectively. The anodal iontophoretic transport of ALA and m-ALA through porcine skin in vitro was followed for 15 h at a constant current of 0.4 mA. When incorporating ALA in the gel, the steady-state was reached in 10 – 12 h at a flux level of approx. 65 nmol cm 2 h 1 compared to 2.5 – 4 h and a level of f 145 nmol cm 2 h 1 for m-ALA. The total amount of m-ALA delivered after 15 h of iontophoresis resulted in a six-fold enhancement over ALA delivery. Iontophoretic delivery from the gel formulation seems to be better than, or comparable to, the passive delivery from formulations commonly used clinically, in spite of the 10 – 20 times lower concentration of the drug in the gel formulation. The skin uptake after iontophoresis for m-ALA showed a nine-fold increase over that of ALA in the stratum corneum (SC). D 2004 Elsevier B.V. All rights reserved. Keywords: Iontophoresis; Aminolevulinic acid; Aminolevulinic acid methyl ester; Carbopol gel; Photodynamic therapy; Skin cancer

1. Introduction Photodynamic therapy (PDT) is a method used for treatment of basal cell carcinoma (BCC), a common type of non-melanoma skin cancer. Irradiation of the tissue at an appropriate wavelength causes a photochemical reaction to occur, resulting in the production of toxic substances, mainly singlet oxygen (1O2), which in turn destroys the tumour cells. 5-Aminolevulinic acid (ALA), a precursor in the heme biosyn* Corresponding author. E-mail address: [email protected] (N. Merclin). 0168-3659/$ - see front matter D 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.jconrel.2004.04.010

thetic pathway, is frequently used in combination with PDT. It stimulates the synthesis of the photosensitizer protoporphyrin IX (PpIX). ALA is applied topically as a prodrug and it accumulates temporarily in cells with an increase in metabolic turnover-rate [1,2]. The penetration depth of ALA is the main factor limiting the efficacy of topical ALA-PDT, a fact that was demonstrated by Fritsch et al. [3] using fluorescence microscopy. Penetration through and into tumours after topical application is possible since ALA is a small molecule. However, the fact that ALA is hydrophilic limits its permeation through biological barriers like cellular membranes or the

58

N. Merclin et al. / Journal of Controlled Release 98 (2004) 57–65

stratum corneum of the skin. In order to enhance dermal bioavailability, a variety of ALA esters has also been investigated [4,5]. With respect to selective accumulation of porphyrins in BCC lesions, the methyl- and hexyl esters are reported to be the best [6]. The formulations commonly used clinically in combination with PDT are water in oil emulsions enriched with 10 – 20% ALA [7,8,9] and, more recently, MetvixR (16% m-ALA). Satisfying results have been obtained for superficial BCCs, but the treatment of nodular or nodularulcerative tumours still presents problems associated with insufficient penetration and the ointments used in vivo are very expensive. Many laboratories have, therefore, investigated the possibility of finding alternative strategies to enable the drug to be delivered more efficiently through and into the tissue. One alternative intended to improve skin penetration of drugs is the use of penetration enhancers. Passive delivery in vivo using DMSO was reported by De Rosa et al. [10]. The use of DMSO increased the flux moderately with a factor 1.6. Another approach is to use iontophoresis, which is a method for transporting charged molecules into and through tissue by application of a small direct current (no more than 0.5 mA cm 2). Iontophoretic investigations have been performed both in vitro and in vivo. Lopez et al. [11] used physiological buffer (pH 7.4) in their in vitro experiments and achieved significant delivery of 15 mM ALA by iontophoresis at pH 7.4. The main mechanism for transport in Iontophoretic delivery is electrical repulsion and electroosmosis. The transport of ALA is mainly due to electroosmosis since only about 10% of the molecule is anionic at physiological pH [12,13]. The iontophoresis of a 2% ALA solution was also tested in vivo by Rhodes et al. [14]. Still, there are many reports concerning the chemical instability of ALA in aqueous solutions, caused by both increasing pH and drug concentration [15,16,17]. Carbopol gels are well documented, and approved for pharmaceutical usage, for several different administration routes. They also represent an interesting media for iontophoresis studies, since iontophoretic devices often utilize hydrogels as a contact interface between the skin and the electrodes [18,19,20]. Cuta-

neous use of these gels is advantageous as they possess good rheological properties, resulting in long residue times at the site of administration. Therefore, they also offer good alternatives to oil based ointment formulations. Carbopol gels too are anionic hydrogels with a good buffering capacity, that may contribute to maintaining the desired pH. The purpose of this study was to evaluate the potential of using a Carbopol gel as a vehicle in iontophoretic administration of ALA and m-ALA, respectively.

2. Materials and methods 2.1. Materials The hydrochloride salts of ALA and m-ALA, 2,4pentanedione, formaldehyde (37%), AgCl (99%) and Ag-wire (99.99%+) were obtained from Sigma Aldrich (Ohio, USA). NaCl (99.9%) was purchased from J.T. Baker (Deventer, Holland), N-2-hydroxyethylpiperazine-NV-2-ethanesulfonic acid (HEPES) from Acros Organics (New Jersey, USA), agar (granulated) from Merck (Darmstadt, Germany) and silicone tubing from Cole Parmer Instrument (Illinois, USA). Poly(acrylic acid) with the proprietary name Carbopol 940 NF (C940) was a gift from Noveon (Brecksville, OH, USA). Ultra-pure water, prepared using a MilliQ water Purification System (Millipore, France), was used in all preparations. Dermatomed pig ear skin ( f 700 Am) was used as the membrane in the in vitro experiments. The tissue was received approx. 2 h after slaughter (Swedish Meats, Uppsala, Sweden) and stored in the freezer (  20 jC) for a maximum period of 1 week before use. 2.2. Preparation of gels Carbopol gels were made by dispersing the polymer powder (1% w/w) in 0.9% NaCl solutions containing the dissolved drug compound (1% w/w). The dispersions were then stirred using magnetic stirring bars for approximately 1 h at room temperature, whereupon 4 M NaOH was added to neutralize each

N. Merclin et al. / Journal of Controlled Release 98 (2004) 57–65

sample to approximately pH 7. All gels were allowed to equilibrate for at least 16 h at room temperature. The pH of the gels was then adjusted to pH 7.3 – 7.5, 0.9% NaCl solution was added to achieve the final volume and the gels were left for at least 90 min before measurement commenced.

59

is the diffusion coefficient of the drug in the gel and t is the time elapsed since the release experiment started. The equation is valid for the first 60% of the fractional release. Plots of the initial drug release versus the square root of time should give a straight line and the diffusion coefficient can be calculated from the slope of that line.

2.3. Rheological measurements 2.5. Preparation of salt bridges The rheological measurements were carried out using a Bohlin VOR Rheometer (Bohlin Reologi, Lund, Sweden), a controlled rate instrument of the couette type. [21]. All measurements were performed at 20 jC, using a concentric cylinder measuring system (C8). Strain sweep measurements were made for all samples to determine the linear viscoelastic region. The elastic (GV) and viscous (GW) moduli were measured using dynamic oscillation, performed within the linear viscoelastic region. 2.4. Determination of diffusion coefficients The drug release from the gels was measured by the USP paddle method with three measurements being conducted on each sample. The gels were put in unidirectional flux containers with a fixed volume of 6 cm3 and a surface area of 21 cm2, covered by a coarse mesh-size plastic net and a stainless steel net. The gel containers were immersed in 300 ml of 0.9% NaCl-solution, stirred at 20 rpm, and maintained at 20 jC using a Pharma Test PTW II USP bath (Pharma Test Apparatebau, Germany). The stirring rate was chosen so that it would give adequate convection and minimize surface erosion of the gels. Samples of 1 ml were taken manually from each gel container and, for each sample taken, the loss of volume was compensated for by immediate addition of the same amount of 0.9% NaCl. The samples were analyzed with HPLC (see Section 2.9 for further details). Under sink conditions during the initial part of the release, one-dimensional Fickian diffusion from a gel holder can be expressed by:  1=2 Dt Q ¼ 2C0 ð1Þ p where Q is the amount of drug released per unit area, C0 is the initial concentration of the drug in the gel, D

To ensure that a constant current was maintained throughout the experiments, salt bridges were used. An agar suspension was prepared as follows: 50 ml of water was heated in an Erlenmeyer flask to boiling. Then, 1.75 g agar was added. Stirring was commenced and the solution was continuously heated until a uniform suspension was formed. At that point, 7.305 g NaCl was added and stirring was continued until the salt had dissolved. The mixture was then cooled under tap water to about 30 jC and transferred to a silicon tube of appropriate length. The tube was then left at room temperature until the mixture inside had gelled [22]. 2.6. Iontophoresis The iontophoresis experiments were carried out in vertical flow-through diffusion cells (Laboratory Glass Apparatus, CA, USA) in a manner similar to that used by Glikfeld et al. [23]. The skin was placed in the cell with the epidermis facing the donor solution. The skin surface area exposed in each electrode chamber was 0.8 cm2. Buffer solutions (25 mM HEPES with 133 mM NaCl, pH 7.4) were then placed in the electrode and receptor chambers, and the system was allowed to equilibrate for 2 h. Then the electrode solutions were replaced with fresh solutions, one of which contained 1 ml Carbopol gel with 60 mM drug. A salt bridge was placed between the anodal formulation and a small container with buffer and the Ag electrode therein (Fig. 1a). The cathode and receptor chambers contained buffer. To avoid air bubbles forming in the flowthrough system, the buffer was degassed prior to use. The receptor phase was perfused at 2 ml/h and a constant current of 0.4 mA was applied for 15 h from a custom-made power supply (made by Roland Lundstro¨m at the Department of Physical Chemistry

60

N. Merclin et al. / Journal of Controlled Release 98 (2004) 57–65

tape (Scotch Magic Tape no. 810, 3M, Saint Paul, USA) 1.9 cm wide and 4 cm long, after 15 tapestrippings, the stratum corneum (SC) had more or less been removed (indicated by the glossy appearance of the exposed surface). The tape strips and the remaining skin were put in separate vials and subsequently immersed in 10 ml methanol and left for 24 and 48 h, respectively, to extract the permeant. The amount of the drugs present was determined by HPLC (see Section 2.9 for further details). 2.8. Passive diffusion through skin The cells used for passive diffusion (Fig. 1b) were also of the vertical flow-through type (Laboratory Glass Apparatus). The skin surface exposed in this equipment is 0.67 cm2, and the volume of the receptor was 3.5 ml (the iontophoretic cells had a receptor volume of 6 ml). The mounting of the skin and the equilibration time was the same as described above. Four measurements were performed on each vehicle and each experiment was conducted over a period of 48 h. Fig. 1. (a) Experimental set-up used for iontophoresis, where (I) is the cathode chamber and (II) the anode chamber. A salt bridge (III) was placed between the anode and a container with buffer in order to ensure that the current was constant throughout the run (15 h). (b) (I) The upper half of the in vitro cell used in iontophoresis seen from below. (II) A view from the side and below of the in vitro cell used in the passive diffusion experiments.

at Umea˚ University, Sweden). The Ag/AgCl-electrodes used in the experiments were prepared in a manner similar to that of Green et al. [24]. The measurements were performed using quadruplicates for each formulation. Fractions were collected automatically every 30 min and analyzed by HPLC (see Section 2.9 for further details). 2.7. Skin uptake After terminating an iontophoretic run, the donor compartment was carefully rinsed in 10 ml of buffer followed by 20 ml of methanol. The skin was removed from the cell and pinned onto a board with the SC face up. The part of the skin that had been exposed to the anode was stripped using a piece of adhesive

2.9. HPLC analysis The quantity of ALA and m-ALA was determined by HPLC with fluorometric detection. Since both substances are non-fluorescent, a derivatization procedure had to be performed before analysis according to the method outlined by Sakai and Morita [25]. A high-performance liquid chromatograph equipped with an autoinjector (from Shimadzu, Kyoto, Japan and Gilson, Middletown, WI, USA, respectively) was used in determining the quantities of the fluorescent products. The sample (20 Al) was injected onto a C18 reversed-phase column (Nucleosil, 125 mm
4.6, particle size 5 Am, Macherey-Nagel, PA, USA). The mobile phase was 50% methanol with 0.1% acetic acid and elution was done at 40 jC at a flow rate of 1 ml/min. The detector wavelength was set at 373/463 nm (excitation/emission). The retention time for ALA was 3.5 min and that for m-ALA 6.7 min. The calibration curve was linear (with r = 0.999) for ALA over the concentration range 0.85 – 31 AM and for m-ALA over the concentration range 30.8 –416 AM. All analyses were carried out within these linear intervals.

N. Merclin et al. / Journal of Controlled Release 98 (2004) 57–65

61

3. Results and discussion 3.1. Characterization of the drug vehicle For the gel to constitute a good vehicle for iontophoresis, the drug compounds must be allowed to move fairly freely through the gel. A determination of the diffusion rate in the gel would indicate whether there are any interactions between the gel and the drug compounds. The diffusion of uncharged substances of low molecular weight through a carbopol gel is similar to the diffusion seen in water. The release of positively charged substances might, on the other hand, be somewhat sustained because of interactions with the negatively charged carboxylic groups of the gel. The release rate of ALA from C940 was rapid; there is no indication of any interaction between the drug compound and the gel. ALA is a zwitterion that is positively charged to only about 10% at pH 7.4. The diffusion coefficient was calculated to be 4.4 F 1.2
10 6 cm2 s 1. The result for m-ALA was different, at pH 7.4 positive charges will prolong the release from the formulations as the drug compound is attracted to the negatively charged carboxylic groups in C940. The release of m-ALA was about 10 times slower than that of ALA, rendering a diffusion coefficient calculated to be 3.08 F 0.7
10 7 cm2 s 1. In addition to allowing the drug compound to leave the vehicle, the gel must also stay more or less intact throughout the iontophoresis. Charged compounds may very well affect the rheology of the carbopol gel, C940, owing to electrostatic interactions or contributions to ionic strength from both the drug compound itself and its counterions; the iontophoresis too may affect the rheology. The rheological properties of the formulations were measured at the start of the iontophoresis experiment. After the iontophoretic experiment had been performed, the gels were taken out and the rheological properties were measured once again. As a reference, a drug-containing preparation of equal age was measured, as was a gel preparation not containing any of the drug compounds. The results from the rheological measurements on C940 containing ALA and m-ALA are shown in Figs. 2 and 3. The gel strength—expressed in terms of the elastic (GV) and the viscous (GW) moduli—of the gel containing ALA only underwent small changes in

Fig. 2. The elastic modulus, GV, (a) and the viscous modulus, GW, (b) on 1% C940 with 5 1% ALA before iontophoresis,
1% ALA used in iontophoresis, D 1% ALA after, but not used in, iontophoresis and w without ALA.

times similar to those allowed for iontophoresis. In the gel containing m-ALA, the change is somewhat more pronounced. After an iontophoretic experiment had been performed on the gels, both the elastic and the viscous moduli had decreased compared to the values for the original gels. The decrease of both GV and GW was comparable for the two systems. A possible explanation for this decrease in the viscoelastic properties could be the increased ionic strength in the gel. In the iontophoretic experiment, ions are transported in the system and a salt bridge is in direct contact with the gel (see Fig. 1). Even though the salt bridge appears to be intact on a macroscopic level, ions will be able to migrate into the gel. An increased ionic strength would screen the charges on the gel, resulting in a reduced repulsion between the polymer

62

N. Merclin et al. / Journal of Controlled Release 98 (2004) 57–65

chains and, therefore, reduction in the viscosity and the elasticity of the gel. According to a rheological definition of a gel, it should have a frequency independent GV that is higher than GW over a large frequency range [26,27]. In that respect, all the formulations studied can be classified as gels. However, gels containing either ALA or mALA will eventually degrade. Three to four days after being prepared, inhomogeneities can be observed with the naked eye and some weeks later the entire gel has turned into a miscolored solution of low viscosity. This does, however, not affect the iontophoresis, as it only endures for about 15 h, apart from which the mixtures would have to be freshly prepared owing to stability problems with the drug compounds. 3.2. Iontophoresis and passive diffusion through skin

Fig. 3. The elastic modulus, GV, (a) and the viscous modulus, GW, (b) on 1% C940 with 5 1% m-ALA before iontophoresis,
1% m-ALA used in iontophoresis, D 1% m-ALA after, but not used in, iontophoresis and w without m-ALA.

The anodal iontophoretic delivery of ALA and mALA from the gel is shown in Fig. 4. The steadystate was attained at different times for the two drugs and the flux level differed. When incorporating ALA in the gel, the steady-state was obtained in 10– 12 h at a flux level of approx. 65 nmol cm 2 h 1 compared to 2.5 –4 h and a level of f 145 nmol cm 2 h 1 for m-ALA. The total amount of m-ALA delivered after 15 h of iontophoresis resulted in a six-fold enhancement over ALA delivery (see Table 1). This can be explained by the different transport mechanisms dominating during iontophoresis as a result of the

Fig. 4. Anodal iontophoretic flux of 1% ALA
and 1% m-ALA 5 from a 1% carbopol gel. Results from iontophoresis of 0.25% ALA in physiological buffer D is also included for comparison (from Ref. [28]). Data shown are the mean F S.D. of four replicates.

N. Merclin et al. / Journal of Controlled Release 98 (2004) 57–65

63

Table 1 Results from anodal iontophoretic delivery of ALA and m-ALA from carbopol gel Drug

Vehicle

Drug concentration (% wt/vol)

pH

Time to steady-state (h)

Flux at steady-state (nmol cm 2 h 1)

Amount of drug passed through the skin after 15 h (nmol cm 2)

ALA m-ALA

carbopol gel carbopol gel

1 1

7 7

10 – 12 2.5 – 4

f 65 f 150

f 356 f 2144

nature of the drug molecule. m-ALA is positively charged at pH 7.4 and will mainly be transported by electromigration. ALA on the other hand will be present as a zwitterion in the formulation for which electroosmotic flow is the main mechanism of transport. Since the amount transported by electromigration is larger than that by electroosmosis for small molecules [13], an increased delivery of m-ALA over ALA was expected. The passive transport experiments where no current was applied showed no detectable amounts of ALA or m-ALA. The delivery of 15 mM ALA (approx. 0.22% wt/ vol) from a physiological buffer (25 mM HEPES with 133 mM NaCl, pH 7.4) had been studied previously in our laboratory [28] and the data from these experiments are included in Fig. 4 and Table 2 for reasons of comparison. Upon adding ALA to the buffer, we observed a drastic lowering of the pH of the buffer and further attempts to adjust the pH to 7.4 were not successful and the final pH of the buffer vehicle was 2 – 3. The gel though has a good buffering capacity and pH f 7 was easily achieved and maintained. The flux from the gel formulation was approximately four times higher than that from the buffer

solution [28]. This difference in flux is, however, not only caused by the four times higher concentration, but also by that different transport mechanisms dominate above or below the pKa of the substance. When comparing the gel formulation to the experiments conducted in buffer solution at pH 7.4 by Lopez et al. [11], the flux is about the same, in spite of the four times lower concentration used by Lopez et al. This might seem surprising, considering the linear concentration dependence for the flux reported by the same authors. However, this may be explained by the flux of ALA being dependent on the ionic strength [12], with a decreased flux at higher ionic strength, since the polyelectrolyte gel contributes somewhat to the ionic strength. The methyl ester of ALA has also been investigated by Lopez et al. [29], who studied the delivery of the drug from an aqueous solution with the pH adjusted to approx. 7 with minimum amounts of diluted NaOH [29]. They achieved remarkable results: After just 2 h a flux of f 4.5 Amol cm 2 h 1 was reached. The high flux was explained by the fact that the anodal solution now contained almost no other competing cations (i.e. the content of Na+ was very low). The lower flux of m-ALA from the gel in our

Table 2 Results from transdermal delivery of ALA and m-ALA found in the literature Drug

Vehicle

Drug concentration (% wt/vol)

Method of delivery

pH

Time to steady-state (h)

Flux at steady-state (nmol cm 2 h 1)

Reference

ALA

0.25

Iontophoresis

2–3

3–4

f 15

[28]

0.22

Iontophoresis

7

[11]

0.25 20

Iontophoresis Passive

f 4–5 f2

f 85 –

[28] [28]

m-ALA

water

0.22

Iontophoresis

7

–

[29]

m-ALA m-ALA

sponge phase Metvix

0.25 16

Iontophoresis Passive

7 f 4–5

steady-state not reached 2.5 – 3 steady-state not reached steady-state not reached 2.5 – 4 f3

–

ALA ALA

buffer (25 mM HEPES, 133 mM NaCl buffer (25 mM HEPES, 133 mM NaCl sponge phase Unguentum Merck

f 80 55

[28] [28]

ALA

64

N. Merclin et al. / Journal of Controlled Release 98 (2004) 57–65

study can be explained by the presence of competing cations as well as the fact that interactions probably arise between the positively charged m-ALA and the negatively charged carboxyl groups in the gel as seen by the low diffusion rate discussed in the previous section. Preparations of 20% ALA in Unguentum M and MetvixR (16% m-ALA) are two commonly used formulations clinically in PDT. Passive diffusion of the drug from these formulations in vitro [28] is given in Table 2. The fluxes are relatively high (but accompanied with large errors) and comparable in magnitude to the iontophoretic fluxes obtained from the gel formulations. Winkler et al. [30] studied the passive flux of ALA from a water in oil formulation (ExcipialR Fettcreme) through human stratum corneum. The passive flux was 2.85
10 11 g cm 2 s 1, which corresponds to 0.61 nmol cm 2 h 1. It seems that iontophoretic delivery from the gel formulation is better than the passive delivery from ointments mentioned above, in spite of 10 –20 times lower concentration of ALA or m-ALA in the gel formulation. A different formulation for the iontophoretic delivery of ALA and m-ALA has also previously been investigated [28]. The vehicle used was a sponge phase consisting of water/propylene glycol/monoolein, in quantities of 41%:29.5%:29.5%. The three components are considered as penetration enhancers on their own. The drug concentration used was 15 mM and the final pH was f 4 – 5 because of difficulty adjusting the pH. As can be seen in Table 2, the sponge phase preparation reached steady-state much sooner and the flux was higher than for ALA in the gel formulation. The difference in pH will not affect the iontophoretic flux to any large extent [11], although the dominant transport mechanism changes from electroosmosis (pH 7.4) to electromigration at lower values of pH. The gel preparation gave a better reproducibility than the sponge phase. The flux from formulations of m-ALA in the sponge phase at pH 7.4 (with same components and in the proportions as mentioned above) was approximately half that of the flux from m-ALA from the gel. The difference in concentration of the drug in the formulations may explain some of the difference. The two preparations took about the same amount of time to reach steady-state.

Table 3 Distribution of ALA and m-ALA respectively into the skin (stratum corneum and viable skin) following 15 h of iontophoresis Drug

Vehicle

pH

Uptake in SC (nmol)

Uptake in viable skin (nmol)

ALA ALAa m-ALA m-ALAa

carbopol gel water carbopol gel water

7 7 7 7

82 6 751 42

75 10 –b 180

Also included are data from the literature for comparison. a From Lopez et al. [29]. Skin uptake was evaluated after 2 h of iontophoresis. b Quantification not possible due to overlapping peaks.

3.3. Skin uptake The skin uptake was studied for both drugs after the iontophoretic experiment and the results are given in Table 3 together with results from Lopez et al. [29]. The uptake in stratum corneum was nine times higher for m-ALA than for ALA, which corresponds well to the seven-fold increase seen by Lopez et al. Despite the fact that Lopez et al. reached a much higher iontophoretic flux of m-ALA in their study, the uptake in the stratum corneum was larger in our study. The skin uptake of ALA was also higher in our study than in that of Lopez et al. even though the iontophoretic fluxes were similar. There are two possible explanations for the difference in uptake between the two studies. Either that Lopez et al. had a lower drug concentration in the donor formulation, and also had not reached the steady-state in their study after 2 h when the skin uptake was evaluated, or that different formulations were used.

Acknowledgements The authors wish to thank Jan Neelissen at AstraZeneca (So¨derta¨lje, Sweden) for lending us the dermatome and Swedish Meats in Uppsala for providing pig ears.

References [1] Q. Peng, T. Warloe, K. Berg, J. Moan, M. Kongshaug, K.-E. Giercksky, J.M. Nesland, 5-Aminolevulinic acid-based photodynamic therapy, Cancer 79 (1997) 2282 – 2308.

N. Merclin et al. / Journal of Controlled Release 98 (2004) 57–65 [2] J.C. Kennedy, R.H. Pottier, D.C. Pross, Photodynamic therapy with endogenous protoporphyrin IX: basic principles and present clinical experience, J. Photochem. Photobiol., B 6 (1990) 143 – 148. [3] C. Fritsch, P. Lehmann, W. Stahl, K.-W. Schulte, E. Blohm, K. Lang, H. Sies, T. Ruzicka, Optimum porphyrin accumulation in epithelial skin tumours and psoriatic lesions after topical application of g-aminolevulinic acid, Br. J. Cancer 79 (1999) 1603 – 1608. [4] J. Kloek, G.M.J. Beijersbergen van Henegouwen, Prodrugs of 5-aminolevulinic acid for photodynamic therapy, Photochem. Photobiol. 64 (1996) 994 – 1000. [5] J.-M. Gaullier, K. Berg, Q. Peng, H. Anholt, P.K. Selbo, L.-W. Ma, J. Moan, Use of 5-aminolevulinic acid esters to improve photodynamic therapy on cells in culture, Cancer Res. 57 (1997) 1471 – 1486. [6] R. Lasse, Photodynamic therapy—basic principles and clinical experience with Metvix PDT. [7] K. Svanberg, T. Andersson, D. Killander, I. Wang, U. Stenram, R. Andersson-Engels, R. Berg, J. Johansson, S. Svanberg, Photodynamic therapy of non-melanoma malignant tumours of the skin using topical y-aminolevulinic acid sensitization and laser irradiation, Br. J. Dermatol. 130 (1994) 743 – 751. [8] R.M. Szeimies, S. Karrer, A. Sauerwald, M. Landthaler, Photodynamic therapy with topical application of 5-aminolevulinic acid in the treatment of keratoses: an initial clinical study, Dermatology 192 (1996) 246 – 251. [9] C. Fritsch, P.M. Becker-Wegerich, H. Menke, T. Ruzicka, G. Goerz, R.R. Olbrisch, Successful surgery of multiple recurrent basal cell carcinomas guided by photodynamic diagnosis, Aesthetic Plast. Surg. 21 (1997) 437 – 439. [10] F.S. De Rosa, J.M. Marchetti, J.A. Thomazini, A.C. Tedesco, M.V.L.B. Bentley, A vehicle for photodynamic therapy of skin cancer: influence of dimethylsulphoxide on 5-aminolevulinic acid in vitro cutaneous permeation and in vivo protoporphyrin IX accumulation determined by confocal microscopy, J. Control. Release 65 (2000) 359 – 366. [11] R.F.V. Lopez, M.V.L.B. Bentley, M.B. Delgado-Charro, R.H. Guy, Iontophoretic delivery of 5-aminolevulinic acid (ALA): effect of pH, Pharm. Res. 18 (2001) 311 – 315. [12] R.F.V. Lopez, M.V.L.B. Bentley, M.B. Delgado-Charro, R.H. Guy, Optimization of aminolevulinic acid delivery by iontophoresis, J. Control. Release 88 (2003) 65 – 70. [13] M.J. Pikal, The role of electroosmotic flow in transdermal iontophoresis, Adv. Drug Deliv. Rev. 46 (2001) 281 – 305. [14] L.E. Rhodes, M.M. Tsoukas, R. Anderson, N. Kollias, Iontophoretic delivery of ALA provides a quantitative model for ALA pharmacokinetics and PpIX phototoxicity in human skin, J. Invest. Dermatol. 108 (1997) 87 – 91. [15] B. Elfsson, I. Wallin, S. Eksborg, K. Rudaeus, A.M. Ros, H.

[16] [17]

[18]

[19]

[20]

[21] [22] [23]

[24]

[25]

[26] [27]

[28]

[29]

[30]

65

Ehrsson, Stability of 5-aminolevulinic acid in aqueous solution, Eur. J. Pharm. Sci. 7 (1998) 87 – 91. A.R. Butler, S. George, The nonenzymatic cyclic dimerisation of 5-aminolevulinic acid, Tetrahedron 48 (1992) 7879 – 7886. G. Novo, G. Huttmann, H. Diddens, Chemical instability of 5aminolevulinic acid used in fluorescence diagnosis of bladder tumours, J. Photochem. Photobiol., B 34 (1996) 143 – 148. J.B. Phipps, R.V. Padmanabhan, G.A. Lattin, Iontophoretic delivery of model inorganic and drug ions, J. Pharm. Sci. 78 (5) (1989) 365 – 369. M.J. Alvarez-Figueroa, J. Blanco Me´ndez, Transdermal delivery of methotrexate: iontophoretic delivery from hydrogels and passive delivery from microemulsions, Int. J. Pharm. 215 (2001) 57 – 65. A.K. Banga, Y.W. Chien, Hydrogel-based iontophoretic delivery devices for transdermal delivery of peptide/protein drugs, Pharm. Res. 10 (1993) 697 – 702. L. Bohlin, New instrumentation for advanced rheometry, Suppl. Rheol. Acta 26 (1988) 161 – 164. G. Christian, Analythical Chemistry, 5th edition, Wiley, USA, 1994. P. Glikfeld, C. Cullander, R.S. Hinz, R.H. Guy, A new system for in vitro studies of iontophoresis, Pharm. Res. 5 (1998) 443 – 446. P.G. Green, R.S. Hinz, C. Cullander, G. Yamane, R.H. Guy, Iontophoretic delivery of amino acids and amino acid derivatives across the skin in vitro, Pharm. Res. 8 (1991) 1113 – 1120. T. Sakai, Y. Morita, y-Aminolevulinic acid in plasma or whole blood as a sensitive indicator of lead effects, and its relation to the other heme-related parameters, Int. Arch. Occup. Environ. Health 68 (1996) 126 – 132. S.B. Ross-Murphy, H. McEvoy, Fundamentals of hydrogels and gelation, Br. Polym. J. 18 (1986) 2 – 7. K. Almdal, J. Dyre, S. Hvidt, O. Kramer, Towards a phenomenological definition of the term ‘gel’, Polym. Gels Netw. 1 (1994) 5 – 17. N. Merclin, J. Bender, E. Sparr, H. Ehrsson, S. Engstro¨m, Transdermal drug delivery by means of a lipid sponge phase—iontophoretic and passive transport in vitro of 5aminolevulinic acid (ALA) and its methyl ester (m-ALA), J. Control Release (submitted). R.F.V. Lopez, M.V.L.B. Bentley, M.B. Delgado-Charro, D. Salomon, H. van den Bergh, N. Lange, R.H. Guy, Enhanced delivery of 5-aminolevulinic acid esters by iontophoresis in vitro, Photochem. Photobiol. 77 (2003) 304 – 308. A. Winkler, C.C. Muller-Goymann, Comparative permeation studies for y-aminolevulinic acid and its n-butyl ester through stratum corneum and artificial skin constructs, Eur. J. Pharm. Biopharm. 53 (2002) 281 – 287.