The effect of water solubility of solutes on their flux through human skin in vitro: an extended Flynn database fitted to the Roberts-Sloan equation

The edited Flynn database (n=62) for determining the effect of the physicochemical properties of solutes on their skin absorption has been extended (n=114) to give a database for which solubilities of the solutes in water, S(AQ), and their maximum fluxes from water through human skin in vitro, J(MAQ), are known or can be calculated. Besides the six major contributors to the original and edited Flynn database, nine more contributors have been included in the extended database to give 15 contributors. As in the edited Flynn database, data for solutes that were significantly ionized or for experiments using different thicknesses of skin were not excluded from the extended database so that the diversity of the original database was maintained. The extended database was fit to five equations where the independent variables were solubility in octanol (S(OCT)), in water (S(AQ)) or molecular weight (MW) and combinations of those three variables; and the dependent variable was J(MAQ). The best fit was obtained from the Roberts-Sloan (RS) equation: logJ(MAQ) = x + ylogS(OCT) + (1 - y)logS(AQ) - z MW, x = -2.574, y = 0.586, z = 0.00440, r(2) = 0.887, S.D. = 0.399, F = 139. This result is comparable to the best fit published using permeability coefficients, P, as the dependent variable, but gives greater insight into the factors affecting permeation. J(MAQ) is more important clinically because it described how much is permeating per unit area and time, while P is in the units of speed (cmh(-1)). Because of the dependence of J(MAQ) on S(AQ), the selection of new drugs with improved topical delivery should include considerations of their S(AQ) in their design.     

Modeling of flux through silicone membranes from water

Do the Roberts-Sloan (RS) or modified Kasting-Smith-Cooper (KSC) equations that provide good fit to data for maximum flux, from water through mouse or human skin also provide a good fit to data for maximum fluxes through silicone membranes (polydimethylsiloxane, PDMS). The maximum fluxes through silicone membranes from water (J(MPAQ)), molecular weights (MW), solubilities in isopropyl myristate (S(IPM)) and water (S(AQ)) of 31 prodrugs and one parent drug have been fitted to the RS equation, which includes a parameter for dependence on S(AQ), and the KSC equation, which does not, to determine which equation gave the better fit. In addition, the J(MPAQ), MW, S(AQ) and solubilities in octanol (S(OCT)) of 26 diverse molecules from other laboratories were collected and fitted to the RS and KSC equations to determine if the choice of lipid parameter (S(IPM) or S(OCT)) had an effect on which equation gave the better fit. RS gave the better fit to the present prodrug database where: logJ(MPAQ)=-2.454+0.716 logS(IPM)+0.284 logS(AQ)+0.00208 MW, r(2)=0.77. RS also gave the better fit to the database from other laboratories where: logJ(MPAQ)=-2.046+0.667 logS(OCT)+0.333 logS(AQ)-0.00374 MW, r(2)=0.878 after four obvious outliers were removed to give n=22. Thus, data for J(MPAQ) can be fitted to the RS equation, which also provides the best fit to maximum flux from water through mouse or human skin and includes a dependence on S(AQ).     

A comparison of the fit of flux through hairless mouse skin from water data to three model equations

Data for the delivery of total species containing parent drugs from water through hairless mouse skin by prodrugs, logJ(MMAQ), has been fitted to the Roberts-Sloan, RS, the Kasting-Smith-Cooper, KSC, and Magnusson-Anissimov-Cross-Roberts, MACR, equations. The RS model which contains a parameter for the dependence of flux on solubility in water, S(AQ), as well as solubility in the lipid isopropyl myristate, S(IPM), gave the best fit: logJ(MMAQ)=-2.30+0.575 logS(IPM)+0.425 logS(AQ)-0.0016MW, r(2)=0.903. The values for the coefficients to the parameters are quite similar to those obtained when the RS model was fit to flux of solutes from water through human skin, logJ(MHAQ). There was no trend in predicting the under or over-performance of prodrugs based on their fit to the RS model and whether they were more or less soluble than their parent drugs. There was an inverse dependence of logJ(MMAQ) on partition coefficients or permeability coefficients similar to that observed for logJ(MHAQ). The similarities in trends for results for logJ(MMAQ) and logJ(MHAQ) suggests that design directives obtained from mouse skin can be extended to design new prodrugs or select new drugs for delivery through human skin.     

Prediction of transdermal flux of prodrugs of 5-fluorouracil, theophylline, and 6-mercaptopurine with a series/parallel model

Multiple regression analysis of fluxes from suspensions in isopropyl myristate (J(M)) as a function of molecular weights (MW) and solubilities in isopropyl myristate (S(IPM)) and water (S(AQ)) were performed on a data set of 41 compounds (n = 41) comprising 39 prodrugs of 5-fluorouracil (5-FU), theophylline (Th), and 6-mercaptopurine (6-MP), in addition to 5-FU and Th, using four models. Two series/parallel models have been developed that allow an aqueous-only path in parallel with a lipid-only path and with a lipid-aqueous series path for the permeation of solutes through skin: log J(M) = log ?1/[1/(aS(LIPID) 10(PhiMW)) + 1/(bS(AQ)/MW(1/2))] + cS(LIPID)10(PhiMW) + dS(AQ)/MW(1/2)? where a, b, c, and d are coefficients for flux through the lipid and aqueous portions of the series path, the lipid-only path, and the aqueous-only path, respectively, and Phi is the dependence of diffusivity in lipid on MW. In the first series/parallel model, S(LIPID) was predicted by S(IPM), and in the second, solvatochromic series/parallel model, S(LIPID) was predicted by S(IPM)(k MW + Omegai) where Omega(i) is the sum of the solvatochromic terms alpha, beta, pi, and R(2), and k is the coefficient for the dependence of partitioning on MW. Using the n = 41 solutions, the coefficients for the aqueous-only path were very small or not different from zero in the two series/parallel models, so only two-path series/parallel models were compared with the solvatochromic and transformed Potts-Guy models where a homogeneous barrier to permeation was assumed. For each model, one compound at a time was omitted from the data set and new parameter estimates were obtained for these 41-1 solutions and used to predict log J(M) for the omitted compound. The average errors of prediction of log J(M) (experimental log J(M) - predicted log J(M)) for the four models were 0.134 for the series/parallel (r(2) = 0.937), 0.127 for the solvatochromic series/parallel (r(2) = 0.967), 0.150 for the solvatochromic (r(2) = 0.950), and 0.134 log units for the transformed Potts-Guy model (r(2) = 0.944). Thus, the solvatochromic series/parallel model provides fit and predictive ability comparable to or slightly superior to previous models that assumed homogeneity of the diffusional barrier to flux in the rate-determining step, provides further theoretical support against the existence of a high capacity aqueous-only path, and provides further insight into the physicochemical properties that should be incorporated into solutes to optimize their flux. Using the solvatochromic series/parallel model, the parameter estimates for the n = 41 solution were used to calculate the flux of each compound through the two paths. For compounds with log partition coefficients (K(IPM:AQ)) of <0.8, permeation was mostly by the lipid-aqueous series path; for compounds with log K(IPM:AQ) >1.0, permeation was mostly by the lipid-only path; the lipid-aqueous series path exhibited the higher carrying capacity. (c) 2000 Wiley-Liss, Inc.     

Topical Delivery of a Model Phenolic Drug: Alkyloxycarbonyl Prodrugs of Acetaminophen

PURPOSE: To determine whether the delivery of a phenolic parent drug by its alkyloxycarbonyl (AOC) prodrugs through hairless mouse skin would show similar dependencies on water and lipid solubilities that similar prodrugs of more polar heterocyclic amide and imide parent drugs have shown. METHODS: Flux through hairless mouse skin from suspensions in isopropyl myristate (J(MIPM)), solubilities in IPM (S(IPM)) and water (S(AQ)), and partition coefficients between isopropyl myristate (IPM) and pH 4.0 buffer (K(IPM:4.0)) were measured for two series of AOC derivatives of acetaminophen (APAP); their solubilities in pH 4.0 buffer (S4.0) were estimated from S(IPM)/K(IPM:4.0). Log J(MIPM) values were calculated from the n = 43 coefficients for the parameters in the transformed Potts-Guy (Roberts-Sloan) equation, and the average error of prediction (delta log J'(IPM)) was calculated. The J(MIPM), S(IPM), S4.0, and molecular weight (MW) data for this series and two other series were combined with the n = 43 database to give a n = 61 database, and new best fit coefficients were determined for the Roberts-Sloan equation: log J(MIPM) = x + y log S(IPM) + (1 - y) log S4.0 - z MW. RESULTS: All of the 4-AOC-APAP derivatives underperformed based on their predicted log J(MIPM) (delta log J'(MIPM) = 0.275 +/- 0.147 log units) and, although the two more water soluble members of this more lipid soluble series were more effective than APAP, they were only marginally so: <2 times. Addition of three new series to the n = 43 database for the Roberts-Sloan equation did not substantially change the coefficients to the parameters: x, y, z, and r2 = -0.322, 0.530, 0.00337 and 0.92, respectively. CONCLUSIONS: The topical delivery of a model phenolic drug by its AOC prodrugs through hairless mouse skin from IPM shows the same dependence on S(IPM), S4.0, and MW as the delivery of polar heterocycles by their similar prodrugs.     

Simple Rules Defining the Potential of Compounds for Transdermal Delivery or Toxicity

PURPOSE: Simple rules based on readily accessible physicochemical properties enable identification of solutes that penetrate skin very slowly or rapidly. METHODS: Literature in vitro maximal flux values (Jmax) across human skin were collected for 87 penetrants. Penetrants were assigned as "good" (Jmax > 10(-5.52) mole x cm(-2) x h(-1)), "bad" (Jmax < 10(-8.84) mole x cm(-2) x h(-1)) or "intermediate" based on mean +/- 1SD. The feasibility of using readily available physicochemical properties, such as molecular weight (MW), melting point (MP, degrees K), octanol-water partition coefficient (K), water solubility (S, molarity), number of atoms available for H-bonding (HB), in assigning solutes was examined. RESULTS: Good penetrants had MW < or = 152, log S > -2.3, HB < or = 5, log K < 2.6, MP < or = 432. Bad penetrants had MW > 213, log S < -1.6, HB > or = 4, log K > 1.2, MP > or = 223. Discriminant analysis using MW, HB, log K correctly assigned 70% of compounds. Individual success rates were good (88%), intermediate (58%), bad (93%). Aqueous Jmax data for 148 test solutes were used for validation. Discriminant analysis assigned 76% of compounds, with individual rates of good (76%), intermediate (67%), and bad (97%). No good penetrants were misclassified as bad or vice versa. CONCLUSIONS: These rules enable rapid screening of potential drug delivery candidates and environmental exposure risks.     

In vitro evaluation of alkylcarbonyloxymethyl (ACOM) ethers as novel prodrugs of phenols for topical delivery: ACOM prodrugs of acetaminophen

The fluxes (J(IPM)) of a series of alkylcarbonyloxymethyl (ACOM) ethers of acetaminophen (APAP) were measured through hairless mouse skin from suspensions of each prodrug in isopropyl myristate (IPM). Solubilities in IPM, estimated solubilities in pH 4.0 buffer (S(4.0)) and flux data for the 4-ACOM-APAP prodrugs were incorporated into the Roberts-Sloan (RS) database to give new estimates for the independent variables of the RS equation: logJ(IPM)=x+ylogS(IPM)+(1-y)logS(4.0)-zM(W). All but one of the 4-ACOM-APAP derivatives hydrolyzed completely on permeation through mouse skin. Three out of the five prodrugs permeated the skin better than APAP, with a maximum fourfold increase in flux. Biphasic solubility - not solubility in a single solvent - was shown to have the greatest impact on flux. A fit of the new n=66 database to the RS equation gave the following values for x, y, z, and r(2): x=-0.545, y=0.511, z=0.00253, r(2)=0.915. These results demonstrate that the topical delivery of a model phenol, acetaminophen, can be improved by transiently masking the phenolic hydroxyl group as an ACOM ether.     

7-alkylcarbonyl and 7-alkyloxycarbonyl prodrugs of theophylline

Five members (6-10) of an homologous series of 7-alkyloxycarbonyltheophylline (7-AOC-Th) and four members (2-5) of a homologous series of 7-alkylcarbonyltheophylline (7-AC-Th) prodrugs have been synthesized by known methods and characterized. All of the members in both series were much more soluble in isopropyl myristate (S(IMP)) (10-200 times) and in each series, at least one member was more soluble in pH 4.0 buffer (S(AQ)) than Th. However, in the 7-AC-Th series, only the acetyl member, 2, which exhibited about 90% of the S(AQ) of Th, was sufficiently stable to be evaluated - it gave four times the flux of Th/IPM (isopropyl myristate). In the 7-AOC-Th series, all the members were sufficiently stable to be evaluated but the member which exhibited the greatest S(AQ), 6 (methyloxycarbonyl), did not exhibit the greatest flux. Instead, 8 (propyloxycarbonyl), which exhibited the second greatest S(AQ) (about 80% of the S(AQ) of Th), but exhibited over ten times the S(IPM) of 6 gave the greatest flux - two times the flux of Th/IPM. Thus, good biphasic solubility was the best predictor of increased flux. All of the prodrugs delivered only Th through the mouse skin. Only 2/IPM actually delivered more Th into the skin than Th/IPM which correlated with its ability to increase the flux of Th through the skin.     

Topical delivery of 5-fluorouracil (5-FU) by 3-alkylcarbonyloxymethyl-5-FU prodrugs

The 3-alkylcarbonyloxymethyl-5-fluorouracil (3-ACOM-5-FU) prodrugs have been characterized by their solubilities in isopropyl myristate (S(IPM)) and partition coefficients between IPM and pH 4.0 buffer (K(IPM:AQ)). Estimated S(AQ) values have been obtained from S(AQ) = S(IPM)/K(IPM:AQ.) The abilities of the prodrugs to deliver total 5-FU species from IPM suspensions through hairless mouse skin (J(i)) have been evaluated in diffusion cell experiments. All of the prodrugs were much more soluble in IPM than 5-FU, and the propionyloxymethyl (C2) member of the series was almost twice as soluble in pH 4.0 buffer. Except for the acetyloxymethyl (C1) member of the series, the 3-ACOM-5-FU prodrugs exhibited greater S(IPM) and S(AQ) values than the corresponding 1-alkylcarbonyloxymethyl (1-ACOM-5-FU) prodrugs. The 3-ACOM-5-FU prodrugs that exhibited greater S(IPM) and S(AQ) values than the 1-ACOM-5-FU prodrugs also exhibited greater J(i) values, except for the C2 member. The C2 member also gave the largest error in predicting J(i) using the transformed Potts-Guy equation. All of the 3-ACOM-5-FU prodrugs delivered more 5-FU as a percentage of J(i) than the 1-ACOM-5-FU prodrugs but were not more effective as a series at targeting dermal as opposed to transdermal delivery.     

Application of the transformed Potts-Guy equation to in vivo human skin data.

Data developed by Wenkers and Lippold for the flux of 10 nonsteroidal anti-inflammatory drugs from light mineral oil (MO) through human skin in vivo has been analyzed using the transformed Potts-Guy equation. The analysis shows that the flux is dependent not only on the solubility in MO (S(MO)), but also on the solubility in acidic water (S(AQ)). This dependence of flux on S(AQ) shows that the previously reported dependence of flux on S(AQ) from in vitro experiments using hairless mouse skin is not an artifact of the in vitro experiments but is due to a characteristic of the skin barrier. Further inspection of the equations used by Wenkers and Lippold in their analyses of their data shows that the equations are variations of the transformed Potts-Guy equation.     

Topical delivery of 5-fluorouracil (5-FU) by 1,3-bisalkylcarbonyl-5-FU prodrugs.

The solubilities in isopropyl myristate (S(IPM)) and pH 4.0 buffer (S(AQ)) and the partition coefficients between IPM and pH 4.0 buffer (K(IPM:AQ)) have been measured for a series of 1,3-bisalkylcarbonyl-5-fluorouracil prodrugs (1,3-AC-5-FU). The 1,3-AC-5-FU prodrugs were each over 500 times more soluble in IPM, but all members of the series, whose solubilities could be estimated, were much less soluble in pH 4.0 buffer than 5-FU. The abilities of the 1,3-AC-5-FU prodrugs to deliver total 5-FU species through hairless mouse skin from IPM suspensions (J(i)) were also measured. The 1,3-diacetyl derivative 2, which exhibited the highest S(AQ) in the series, gave the highest J(i) value. Although the series of 1,3-AC-5-FU prodrugs was generally effective at increasing J(i) (three to ten times), the best 1,3-AC-5-FU prodrug was not as effective as the best 1- or 3-alkylcarbonyl-5-FU prodrug (1- or 3-AC-5-FU) at increasing J(i) and their ability to increase the concentration of total 5-FU species in the skin was generally less than that of the 1-AC-5-FU prodrugs, but greater than that of the 3-AC-5-FU prodrugs. Thus, the 1-AC-5-FU prodrugs remain the best prodrugs with which to enhance the topical delivery of 5-FU.     

Topical delivery of N-alkyl-N-alkyloxycarbonylaminomethyl (NANAOCAM) prodrugs of theophylline (ThH)

N(7)-(N-Alkyl-N-alkyloxycarbonyl) aminomethyl (NANAOCAM) prodrugs of theophylline (ThH) have been synthesized and characterized by their solubilities in isopropyl myristate (S(IPM)), solubilities in water (S(AQ)), partition coefficients between IPM and pH 4.0 buffer (K(IPM:4.0)) and by their ability to penetrate hairless mouse skin from IPM (J(MIPM)). The most lipid soluble and water soluble member, N-methyl-N-ethyloxy-carbonylaminomethyltheophylline, gave the highest flux through hairless mouse skin from IPM compared to ThH. The flux of NANAOCAM prodrugs of ThH can be accurately predicted by the Roberts-Sloan (RS) equation.     

Estimation of the ethanol/water solubility profile from the octanol/water partition coefficient

While the ethanol/water solubility profiles of very polar and very non-polar drugs are monotonic, many semi-polar drugs show a maximum solubility at an ethanol volume fraction (f(max)) between 0 and 1. A sigmoidal relationship was observed between the value of f(max) and the log of the octanol/water partition coefficient (logK(ow)) of the solute. This relationship reasonably predicts the value of the volume fraction of ethanol that gives maximum solubility (f(max)). Combining this sigmoidal relationship with the previously reported linear relationship between the logK(ow) and the initial slope of the plot of log solubility versus ethanol composition [Li, A., Yalkowsky, S.H., 1994. Solubility of organic solutes in ethanol/water mixtures. J. Pharm. Sci. 83, 1735-1740] enables the estimation of the total ethanol/water solubility profile.     

Lipid extraction and transport of hydrophilic solutes through porcine epidermis

The purpose of this study was to investigate the effect of delipidization of the stratum corneum (SC) on the in vitro percutaneous absorption of hydrophilic solutes (i.e. water, urea, and inulin). Fourier transform infrared (FT-IR) spectroscopy was employed to study the extent of delipidization of porcine SC due to chloroform:methanol (2:1) (C:M (2:1)) treatments for various time periods. In vitro percutaneous absorption of [3H] water, [14C] urea, and [3H] inulin were studied through C:M (2:1) treated epidermis in Franz diffusion cells. There was a greater decrease in peak areas of the asymmetric and symmetric C-H stretching absorbances (i.e. increase in lipid extraction) with increasing exposure times of the SC with C:M (2:1). After 40-min treatment, asymmetric and symmetric C-H stretching peak area showed a decrease of 75.9 and 89.9%, respectively. The permeability coefficient of water, urea, and inulin increased with increasing lipid extraction. Enhancement in the permeability coefficient, through 40 min C:M (2:1) treated epidermis in comparison to the control, for water, urea, and inulin was 48.72, 215.65, and 3.90, respectively. Log (permeability coefficient) and log (mol. wt.) for test solutes and leuprolide acetate were found to be inversely related (R(2)=0.9974). In conclusion, this study implies that penetration enhancers that are safe and extract the SC lipids can be selected in order to enhance the percutaneous absorption of polar solutes through the skin.     

Kinetics of talampicillin decomposition in solutions

Talampicillin stability in aqueous solutions was studied in a broad range of pH values using as medium solutions of hydrochloric acid (pH 0.4-1.8), phosphate buffers (pH 2.05-3.13 and 6.03-8.04), acetate buffer (pH 3.87-5.28) and borate buffer (pH 8.90-9.10) as well as sodium hydroxide solution (pH 11.48). For the determination of talampicillin concentration changes in kinetic studies, two methods were used: iodometric and spectrophotometric in UV (lambda(max) = 254.5 nm). The catalytic velocity constants (k(H+), k(x), k(o)) were established, the log k-pH profile (35 degrees C) was found, thermodynamic parameters were calculated of the hydrolysis reaction of the beta-lactam bond (k(H+): E(A)= 67.9 kJ mol(-1), delta S = -92.4 J K(-1) mol(-1), delta G = 92.6 kJ mol(-1); k(x): E(A) = 31.8 kJ mol(-1), delta S = -347.1 J K(-1) mol(-1), delta G = 131.1 kJ mol(-1); k(o), pH = 5.28: E(A) = 98.0 kJ mol(-1), delta S = -50.3 J K(-1) mol(-1), delta G = 110.3 kJ mol(-1) at 20 degrees C), and the stability of the lactone bond was studied in the medium with the highest stability of beta-lactam bond of talampicillin (pH 5.28: k(o): E(A)= 32.5 kJ mol(-1), delta S = -220.5 J K(-1) mol(-1), delta G = 94.7 kJ mol(-1) in 20 degrees C), at controlled ionic strength (mu = 0.5 mol l(-1)).     

Mechanistic analysis of pH-dependent solubility and trans-membrane permeability of amphoteric compounds: application to sildenafil

This study was aimed at investigating the pH-dependent solubility and in vitro transmucosal permeability of sildenafil, an amphoteric compound with limited aqueous solubility, across parallel artificial membrane. The aqueous solubility and permeability of sildenafil as a function of solution pH were theoretically derived from the individual contributions of all species (cationic, neutral and anionic). The stability, octanol-water distribution coefficient (log D), and solubility of sildenafil were then determined at various pHs, the permeability study was also performed at different pHs using parallel artificial membrane. The pH-solubility and -permeability profiles were then fitted to theoretical equations using non-linear regression. The experimental pH-solubility profile was fitted very well to the theoretical equations (R(2)=0.9996). The in vitro permeability of saturated sildenafil solution at different pH values also showed similar trend as the predicted one (R(2)=0.7829). The two optimum pH (pH(max)) values were found to be 4.50 and 10.24, where the maximum solubility of either cationic or neutral species, or anionic and neutral species is simultaneously obtained, and the maximal transmucosal fluxes (J(ss)) are achieved. The above method can be applied to optimize the transmucosal delivery of other amphoteric drugs with low aqueous solubility.     

Afloqualone phototoxicity--photosensitized lipid peroxidative potency and its biological significance

Photosensitivity due to afloqualone (AQ), a muscle relaxant acting on the central nervous system, has tended to increase with increasing clinical use of this drug by orthopedists. The initial step leading to drug photosensitivity is a photochemical reaction between photo excited phototoxic compounds (drugs) or their metabolites and target molecules in skin. The present study was an attempt to determine whether unsaturated lipids are one of the target molecules of AQ-photosensitivity. Squalene, an unsatulated model lipid, was subjected to UVA irradiation in the presence of AQ and production of squalene hydroperoxides was confirmed. AQ-photosensitized peroxidation of squalene was repressed in the presence of 2.5-dimethylfuran, a singlet oxygen acceptor. Production of singlet oxygen by UVA irradiation of AQ was also observed in aqueous solution. These findings suggested participation of singlet oxygen in the AQ photosensitized peroxidation of squalene. Although photohemolysis was not observed by UVA irradiation in the presence of AQ, We propose that lipids can be a target molecule of AQ-phototoxicity, and lipid peroxidation might be one of the important factors responsible for induction of the AQ phototoxic reaction.