An in vitro Skin Penetration Model for Compromised Skin: Estimating Penetration of Polyethylene Glycol [¹⁴C]-PEG-7 Phosphate

A good understanding of systemic bioavailability of ingredients is essential to predict the exposure to consumer product ingredients that are applied topically. Stratum corneum (SC), the top layer of the epidermis is the predominant barrier for permeability, protecting against infections, preventing chemical penetration, supporting thermoregulation and hydration [1,2]. Intact skin in healthy full-term infants exhibits skin barrier properties comparable to those of adults based on transepidermal water loss (TEWL) measurements [3]. In contrast, premature infant skin is known for thinner SC and underdeveloped epidermis/dermis, resulting in a decreased barrier function with higher water loss [1,3,4,5,6,7,8]. To evaluate the potential for higher absorption from topically applied products through compromised skin or skin with underdeveloped barrier function as in premature infant skin, modifications of existing in vitro skin penetration protocols are needed. The objective of this study was twofold: (1) to establish an in vitro skin penetration model for moderately and highly compromised skin, by progressive tape stripping intact skin to physically impair skin barrier function; (2) to determine differences in dermal penetration of a topically applied substance in multiple-dose versus single-dose application scenarios equivalent to a cumulative 24-hour multiple-dose exposure, mimicking consumer products used once or multiple times a day.

The test material consisted of sodium PEG-8 phosphate in a typical wet wipe lotion. It is a polydisperse mixture of polyethylene glycol (PEG) monophosphate, diphosphate, and some nonphosphated species of varied chain length (Rhodia, UK). PEG-7, PEG-9 and PEG-11 were the most prevalent species in the mixture. Radiolabeled PEG-7 monophosphate (molecular weight, MW approx. 417 Da) was used as a conservative surrogate for the overall PEG-8 phosphate raw material to simplify the radiolabeling process and provide a conservative estimate for potential skin penetration (radiolabeled at Procter & Gamble, Cincinnati, Ohio, USA). The test material was produced at a final PEG-8 phosphate surrogate material concentration of 2% (w/w). The chemical authenticity of the surrogate [¹⁴C]-PEG-7 monophosphate was confirmed by HPLC and mass spectrometry. Any detected counts were reflective of the overall penetration of the radiolabeled surrogate. The test material sodium PEG-7 phosphate was chosen as a model compound with moderately low MW and hydrophilic properties representative of ingredients in baby wipe lotion.

Human skin samples were obtained from 5 donors (plastic surgery patients). Each treatment group [intact, moderately and highly compromised skin (repeat dose); intact and highly compromised (single dose)] had a total of 12 human skin samples per group from 5 different donors. The samples from each of the 5 donors were randomly selected (n = 2-3) to randomize any variability due to donor skin differences for best comparison of results between groups and in accordance with SCCS guidelines [5]¹. Split-thickness membranes of 400 µm were prepared using an electric dermatome (Zimmer, Ltd., UK). Samples were stored at -20°C prior to experimentation. Ethical approval for the use of human skin was obtained.

The degree of skin compromise was evaluated by TEWL, which is extensively used to characterize skin barrier function. TEWL measures the steady-state water flux that passes through the epidermal layer to the surrounding atmosphere via diffusion and evaporation [9,10]. TEWL has been used as an indicator of skin integrity deficiencies, like damage from chemical or physical irritants or changes under occlusive conditions [11,12,13,14].

Prework to Determine Tape Strips Needed to Compromise Skin. The barrier function from 5 human skin donors was impaired by tape stripping. The TEWL measurements were taken in the following groups (n = 3): intact, 5, 10, 20, and 30 tape strips (Scotch Magic tape 3M France, Cergy-Pontoise, France). A control sample from each of the 5 donors was prepared by heat separation of the epidermis from the dermis. This provided a TEWL measurement from each donor that had no SC barrier present. Skin barrier integrity was evaluated by measuring TEWL using AquaFlux AF200 (Biox Systems, Ltd., UK). The results obtained determined the number of tape strips required to obtain moderately compromised (SC removed with 20 adhesive tape strips) and highly compromised (SC removed with 25-30 adhesive tape strips) barrier function (table 1). Impairment of skin integrity was also confirmed by examining histological samples of skin with SC removed at various levels of tape stripping (fig. 1).

The study was conducted at the Charles River Laboratories (Tranent, UK) following the OECD Principles of Good Laboratory Practice and specified modifications related to creating compromised skin samples [15]. Test material was applied to skin samples in a quantity to simulate actual usage conditions of baby wipe lotion and was covered with semiocclusive diaper backsheet material which was placed directly on top of the samples. Parafilm® Wrap (Cole-Parmer, USA) was used to hold the backsheet in place. The skin samples were mounted in an automated flow-through diffusion cell system (Scott/Dick, University of Newcastle-upon-Tyne, UK), maintaining a skin surface temperature of 32 + 1°C. The skin samples were dosed topically with [¹⁴C]-PEG-7 phosphate (PEG-8 phosphate surrogate material) incorporated into a baby wipe lotion formulation in two exposure scenarios: (i) a repeat dose of 5 mg/cm² applied to intact, moderately and highly compromised skin for a total of 5 applications and wash cycles over a 24-hour period (groups 1, 2 and 3, respectively), and (ii) a single dose of 25 mg/cm² applied to intact and highly compromised skin samples (groups 4 and 5, respectively) over a 24-hour period (table 2). In both application scenarios a total overall dose of 25 mg/cm² was applied. The test material was dried for 2 min prior to being covered with semiocclusive diaper backsheet material to mimic use conditions. At 3 h 45 min skin samples from groups 1-3 were washed twice with phosphate-buffered saline (pH 7.3-7.4), twice with water, and dried with tissue swabs prior to repeat application. For all groups (groups 1-5), the final wash after 24-hour application was with water. The repeat dosing/washing was successfully validated in this laboratory [16]. The skin samples were dried with tissue swabs. The tissue swabs and wash fluids in all application scenarios were retained for analysis. The receptor fluid (degassed phosphate-buffered saline, pH 7.3-7.4; Sigma, St. Louis, Mo., USA) was collected hourly from 0 to 24 h after initial application. After final washing and drying, skin samples were removed from the cells and SC was removed with 35 tape strips for intact skin samples. For moderately and highly compromised skin samples the SC was removed using up to 35 tape strips in total, including pre-tape stripping that was used to initially compromise skin. The epidermis was separated from the dermis of the exposed skin. The epidermis, dermis, and unexposed skin from all samples were analyzed for [¹⁴C]-PEG-7 phosphate by liquid scintillation counting (Packard Scintillation Analyzer Model 2100TR).

The absorbed dose was evaluated from the hourly fractions of receptor fluid, receptor rinse and receptor chamber wash fluid. Absorption was expressed as microgram equivalents per square centimeter, and as percent of applied dose. The total bioavailability (dermal penetration) was the sum of applied dose found in the treated skin (epidermis and dermis) and absorbed dose at the end of the experiment.

The total bioavailability (dermal penetration, μg Eq/cm²) was calculated as the sum of applied dose found in the treated skin (epidermis and dermis) and absorbed dose (receptor fluid, receptor rinse and receptor chamber wash fluid) at the end of the experiment. A statistical comparison of dermal penetration of radiolabeled PEG-7 phosphate through intact, moderately, and highly compromised skin samples was performed. Data was analyzed using a one-way analysis of variance. All statistical tests were two-tailed and performed at the 5% significance level using SAS v8.2 (p < 0.05).

Skin samples were compromised by tape stripping intact skin and skin barrier integrity was determined by TEWL. The number of tape strips needed for a representative moderately and highly compromised skin was determined during pretest of donor samples (table 1). TEWL values obtained from samples are also listed in table 1. Twenty tape strips for moderately compromised and 25-30 tape strips for highly compromised skin was sufficient to achieve the desired compromise level/TEWL based on literature review [17]. Samples with heat-separated epidermis were used to represent the maximum possible damage to the skin barrier and maximum achievable TEWL (table 1).

The test material application regimen for repeat and single dose, and the 24-hour cumulative total absorption, mass balance, and total bioavailability of the test material are presented in tables 2 and 3, respectively. A total of 12 human skin samples per group (groups 1-5) was randomly selected from 5 different donors (n = 2-3), in accordance with SCCS guidelines [5].

Intact skin exposed to repeat (5 × 5 mg/cm², group 1) and single (25 mg/cm², group 4) dose of [¹⁴C]-PEG-7 phosphate for 24 h showed a total penetration (material in receptor fluid, epidermis, and dermis) of 4.18 ± 1.32 µg Eq/cm² (0.76 ± 0.24% of applied dose), and 6.03 ± 2.14 µg Eq/cm² (1.27 ± 0.45% of applied dose), respectively (fig. 2a, b, 3a, b, table 3). Moderately compromised skin (group 2) exposed to repeat dose of [¹⁴C]-PEG-7 phosphate showed a directional increase in absorption after each fresh application of material. The total penetration was 24.66 ± 18.82 µg Eq/cm² (4.48 ± 3.42% of applied dose; fig. 2a, 3a, table 3). Single-dose exposure of 25 mg/cm² was not conducted for moderately compromised skin. Highly compromised skin (group 3) exposed to repeat dose of [¹⁴C]-PEG-7 phosphate showed an increase in absorption with each fresh application of material. The total penetration was 17.64 ± 12.58 µg Eq/cm² (3.19 ± 2.06% of applied dose; fig. 2a, 3a, table 3). Single-dose exposure of 25 mg/cm² on highly compromised skin (group 5) indicated absorption that leveled after 4-5 h resulting in an overall 24-hour cumulative penetration of 20.86 ± 10.26 µg Eq/cm² (4.42 ± 2.17% of applied dose), compared to intact skin (group 5; fig. 2b, 3b, table 3).

In the repeated dosing conditions the total penetration through moderately and highly compromised skin (groups 2 and 3) was significantly higher versus intact skin (group 1, absorbed dose and total bioavailability: p < 0.001; fig. 2a, table 4). However, there was no significant difference in cumulative penetration between the two compromised groups themselves (group 2 vs. 3, absorbed dose: p = 0.30, total bioavailability: p = 0.23). Single dose of 25 mg/cm² showed a significantly higher cumulative penetration through highly compromised skin compared to intact skin (group 4 vs. 5, absorbed dose: p = 0.001, and total bioavailability: p < 0.001, respectively; fig. 2b, table 4). There was no significant difference between intact skin exposed to a single large dose and intact skin exposed to the repeat dose over 24 h (group 1 vs. 4; absorbed dose: p = 0.45, total bioavailability: p = 0.10). Interestingly, the highly compromised skin samples (groups 3 and 5) also did not show any significant difference in penetration between repeat dose and single large dose (absorbed dose: p = 0.57, total bioavailability: p = 0.13; table 4).

In this study the skin barrier function of normal adult skin was compromised by tape stripping as a model for premature and compromised infant skin. Mechanical damage removes most of the SC, and a correlation between the TEWL and penetration of hydrophilic compounds can be demonstrated. Nine investigations correlating TEWL and percutaneous absorption have been reported, where specific test protocols including healthy and damaged skin, skin from humans and animals, and measures of percutaneous absorption both in vivo and in vitro were used [14]. In most of these studies a significant quantitative correlation between TEWL and percutaneous absorption was demonstrated. The TEWL of full-term neonates has been reported to be <10 g/m²/h, similar to that of healthy adults [3,18]. However, preterm infants have an immature epidermal barrier with a much thinner SC and higher TEWL as reported by Agren et al. [19] in a study of 13 preterm infants born at 24 and 25 weeks of gestation. Mean TEWL values at 1 day after birth were 58.4 ± 14.8 g/m²/h. At 28 days after birth, TEWL values had decreased to 24.2 ± 7.7 g/m²/h, but were still elevated above what is considered normal for full-term infants [19]. Kalia et al. [7] reported a study of 10 neonates that were born at 30 weeks or less of gestational age with elevated TEWL levels at birth. Infants born at >30 weeks gestational age showed TEWL values comparable to normal adult skin (i.e. <10 g/m²/h) [7]. Several other studies conducted on preterm infants showed an increased TEWL for infants born at <30 weeks gestational age [6]. Consequently, use of appropriate TEWL to determine skin barrier function was an integral part in the establishment of this skin penetration model and therefore closest to mimic the skin conditions of a premature infant for hydrophilic compounds like PEG.

In this study, the TEWL and tape strip cutoffs for in vitro modeling of compromised skin barrier such as premature skin could not be directly derived from published in vivo premature TEWL values, which are often measured by open-chamber devices. Extrapolation of the published in vivo premature TEWL data to our in vitro experimental settings was made possible considering the following points: (i) Increased diffusion resistance and reduced flux (TEWL) for in vitro setting compared to in vivo measurements on premature infant skin (which is a low-resistance skin barrier) by AquaFlux (due to the increased distance between skin and condenser 16 vs. 28 mm). We estimated that this increased distance makes the in vitro AquaFlux measurements (which are often greater than the open-chamber measured values) comparable to the readings obtained from other open- and closed-chamber TEWL instruments like Servomed. (ii) The inverse correlation of membrane thickness and TEWL (according to Fick's first law of diffusion). With this consideration, we tried to find the optimum number of tape strips that would result in a skin thickness for which the TEWL would match the published TEWL values from premature infants. Correlating TEWL with SC thickness, Sekkat et al. [17] presented a model to predict the barrier development level of a premature neonate based on its postconceptional age. Superposition of the in vivo TEWL data for premature infants, gestational age, and SC thickness in the model of Sekkat et al. [17] revealed that (a) for moderately compromised skin, premature infants at 29-31 weeks of gestational age possess 40-60% of full SC thickness and the corresponding TEWL was in a range of 10-20 g/m²/h. (b) For highly compromised skin, TEWL measurements corresponding to SC thickness are estimated to be 30-35 g/m²/h. These TEWL ranges matched the average TEWL data obtained from 5 skin donors in the pretest after 20 (for moderately compromised) and 25-30 (for highly compromised) tape strips on intact skin samples. However, given inherent interindividual variability in SC thickness and corneocyte cohesion, the target TEWL range was not expected to be achievable for all 5 donors upon the same number of tape stripping.

Based on comparable TEWL values, it is well known in the scientific literature that full-term infant skin barrier is similar to that observed in adult intact skin [3,4,5,18]. Several methods have been used to quantify the functional maturity of the skin barrier in term and preterm infants, including measurements of transdermal fluid loss, carbon dioxide emission rates, and drug absorption. In general, these have supported the idea that intact skin in healthy term and near-term infants exhibits barrier function comparable to older children and adults [3]. Using TEWL as the gold standard, the literature evidence shows that full-term neonates have completely developed permeability barrier function at birth. High TEWL was shown in the first 4 h after birth, while TEWL afterwards returned to values of lower than 10 g/m²/h (normal range for healthy skin under basal conditions), suggesting ongoing drying of the skin immediately after birth [3,4,9,20]. However, depending on the exposure conditions (open to air, semiocclusive, occlusive) and the physicochemical properties of the permeants and the composition of the vehicle, dermal penetration and systemic bioavailability may be increased for compromised skin conditions such as premature infant skin, or in diaper dermatitis areas. Skin barrier can be compromised by different factors such as solvents, surfactants, hydration and mechanical disruption, e.g. by tape stripping. It has been shown that the penetration enhancement by such factors ranges from 1.5 to 1,300 [21]. Gattu and Maibach [22 ]demonstrated for small-MW compounds (MW ≤1,000), that skin penetration enhancement through compromised human skin by mechanical (tape stripping or abrasion) or chemical damage compared to intact skin is higher for hydrophilic versus lipophilic chemicals in in vitro studies. Further, increased penetration was demonstrated for autopsy premature skin when compared to adult and full-term infant (38-40 weeks gestation) skin samples [23]. The degree of impact of damaged skin on dermal penetration rate is chemical-specific, with some chemicals having an increased and others having a comparable skin penetration to that of intact skin [11,12,22]. Substances with MW >500 Da are generally accepted to penetrate slowly through intact skin with negligible penetration for molecules with a MW >1,000 Da in intact adult skin [5]. In general, the systemic availability of higher-MW compounds is lower following dermal exposure [24,25]. However, increased dermal penetration irrespective of MW has been demonstrated for compromised skin conditions [24]. Several PEGs with a MW from near 300 to over 1,000 Da have been tested in in vitro penetration studies using different degrees of skin damage (TEWL: 10-40 g/m²/h) and different mechanisms (tape stripping, sodium dodecyl sulfate) to mimic compromised skin conditions [11]. It was demonstrated that a 12-hour penetration of all PEG oligomers (MW 300-1,000) generally increased with increase in TEWL when skin was compromised either by tape stripping or sodium dodecyl sulfate pretreatment. All PEG 600 oligomers penetrated compromised skin at TEWL >13 g/m²/h, while the PEG 1,000 oligomers were found only at TEWL >20 g/m²/h. There was no significant difference between the two methods (sodium dodecyl sulfate vs. tape strip) to compromise the skin. The total amount penetrated was approximately 1% of the applied dose for the high-MW oligomer (MW 1,074 Da; TEWL: 30-40 g/m²/h) and approximately 7% for the smallest PEG (MW 238 Da; TEWL: 30-40 g/m²/h) used in the test [11]. For small-MW PEGs (282-590 Da) Jakasa et al. [12] demonstrated a threefold increase in the permeability coefficient in SLS-treated skin [5% sodium lauryl sulfate for 4 h (forearm skin), and 24 h later patched with PEGs for 6 h] compared to untreated skin. The permeability coefficient for all MWs in untreated skin ranged from 0.34 to 0.70 × 10^-5 cm/h and in sodium lauryl sulfate-damaged skin ranged from 1.20 to 2.09 × 10^-5 cm/h for MW of 590-282 Da [12]. Thus, even compounds having a high MW can have a potentially higher bioavailability when exposed to compromised versus intact skin.

In our investigation, diffusion of the test substance [¹⁴C]-PEG-7 phosphate through intact skin resulted in lower penetration of test material compared to damaged skin for both repeat-dose and single-dose applications (table 3). These penetration amounts are relatively small compared to the total dose applied. Skin penetration of [¹⁴C]-PEG-7 phosphate even via compromised skin was low (<5%) compared to intact skin. This finding is consistent with other publications, indicating that there may be a modest enhancement of penetration for compromised skin compared to intact skin [11,22,23]. However, sodium PEG-8P, the active substance in the wipe lotion, is a charged molecule and is anticipated to have even lower SC penetration potential than the surrogate test material [¹⁴C]-PEG-7 phosphate. The skin permeability coefficient of molecules with ionized fractions >60% is known to be between 0.5 and 2.5 orders of magnitude lower than that of mostly nonionized compounds [26]. It is anticipated that SC will provide a high diffusive resistance toward ionized chemicals, such that a high percentage of the applied dose will not penetrate into the deeper tissue layers.

Our model of using different degrees of compromised skin in an in vitro setting enables an understanding of the rate of dermal penetration of materials intended for topical use by consumers with intact, compromised, or immature skin barrier. Our findings indicate that: (i) it is possible to compromise skin using an in vitro model based on tape stripping and increased TEWL that may represent compromised skin such as that of premature infants. (ii) Even under highly compromised skin conditions, penetration of [¹⁴C]-PEG-7 phosphate through skin is low (<5%) and only 4-6 times higher compared to mature/intact skin. (iii) Dermal penetration of [¹⁴C]-PEG-7 phosphate is not affected by single or multiple dosing under present study conditions. (iv) Even under highly compromised skin conditions, skin penetration is low and clearly does not approach 100%.

Our data supports that even under compromised skin conditions it is not appropriate to assume a general high level of absorption. Additional data/experimentation is necessary to understand the relationship of skin damage to dermal penetration for chemicals with diverse characteristics. Our data indicates that in vitro refinement of dermal penetration of chemicals through damaged skin is feasible using this skin penetration model. In addition, for future dermal penetration studies of chemicals with similar physicochemical properties as PEG-7 phosphate, using a single large dose equivalent to the cumulative 24-hour multiple dose exposure may be an appropriate and somewhat conservative approach to determine dermal penetration in both intact and compromised skin and would have the added benefit of reducing the complexity of future studies.

The authors acknowledge Prof. Bob Imhof, London South Bank University and Biox Systems Ltd. for his technical expertise in TEWL measurements. The authors also acknowledge Ken Yelm for radiolabeling the test material.

Some authors are employees of The Procter & Gamble Company.