The effect of dilution of fusidic acid cream and betamethasone dipropionate cream in complex extemporaneous mixes on formulation performance

The Aron regimen is an unconventional therapy which entails frequent applications of an extemporaneously prepared three component system (a topical antibiotic, a corticosteroid and an emollient), with the intention of decolonising the skin of S. aureus whilst treating atopic dermatitis. The impact of heavily diluting these topical medicinal products, to differing extents, on formulation performance is not well understood thus was investigated in this study. Following a single application of a range of compounded Aron mixes (fusidic acid and betamethasone dipropionate diluted to varying extents in an emollient base), significant reductions in the expected drug flux across silicone membrane, ex vivo percutaneous absorption and skin retention of both drugs relative to the marketed products were observed. This was attributed to a number of complex formulation effects making such changes difficult to predict in a clinical setting. Further investigations are required to evaluate the impact of frequent applications of the Aron mix to widespread areas on clinical efficacy, antimicrobial resistance and long term side effects. Graphical abstract


Introduction
In severe cases of atopic dermatitis patients are prescribed a combination of therapies, including emollients, topical corticosteroids (TCSs), topical calcineurin inhibitors, systemic treatments and phototherapy. In cases unresponsive to these treatments, patients may resort to unconventional treatments to manage the skin condition. Once such emerging treatment is the Aron regimen, pioneered by Dr Richard Aron, which combines three components of conventional treatment, an emollient, a topical corticosteroid and a topical antibiotic and uses them in an unconventional way. The Aron regime's aim is to address skin colonisation of S. aureus, whilst treating atopic dermatitis (Aron, 2022a). Recent work has reported that S. aureus is extensively present in areas of affected and unaffected skin in up to 90 % of atopic dermatitis patients (Wollenberg et al., 2018), which may have a causative role in inducing atopic dermatitis flares. The Aron regimen compounds the emollient, TCS and topical antibiotic into a single formulation (the Aron mix), tailored to the patient, which is applied to all affected and unaffected areas of the body frequently (up to six times a day for one to two weeks, tapered thereafter according to response to treatment).
Whilst the success of the Aron regimen has been reported by patients, carers and Dr Aron (The Daily Telegraph, 2014;The Guardian, 2018;Aron, 2022a), the body of evidence is largely anecdotal. Case series studies (non-blinded, non-controlled) have reported improvements in atopic dermatitis severity following treatment with a compounded antibacterial, steroid, and moisturizer, however, no controlled clinical studies evaluating the effectiveness of the Aron regime have been conducted to date (Lakhani, Lee and Lio, 2017;Rajkumar et al., 2021). Indeed, recent evidence has reported the limited benefit of a topical antibiotic plus a TCS compared to a TCS alone in the treatment of secondary bacterial infection of dermatitis (George et al., 2019;NICE, 2021). Other criticisms have been made of the approach, including that it may promote the development of antibiotic resistance. One of the cornerstones of the Aron regimen is to tailor the degree of dilution of the TCS and antibiotic depending on the patient age, weight and severity of the condition. The rationale behind heavily diluting the TCS and topical antibiotic is to allow uninterrupted therapy (and more frequent applications) thereby preventing the risk of 'steroid rebound' or recolonisation of the skin by S. aureus. However, the extent to which drug delivery to the skin is altered does not always correlate with the degree of dilution of a product, thus cannot be readily predicted in a clinical setting (Ryatt et al., 1982;Gibson et al., 1983). This informs the recommendation that the extemporaneous dilution of topical products to tailor potency should be avoided (British National Formulary, 2022).
The impact of varying the degree of dilution of an antibiotic and a TCS in an emollient base on the formulation performance is not well understood. Recent work has demonstrated that mixing two topical products can induce complex formulation changes which alters the drug delivery profile relative to the individual marketed products (Beebeejaun et al., 2019). These considerations are equally applicable to a further complex extemporaneous mix of a TCS, antibiotic and emollient. Thus, the aim of this work was to evaluate the impact of varying the degree of dilution of a topical antibiotic and TCS in the Aron mix on in vitro formulation performance compared to the relative marketed products.

Raman microscopy of Diprosone cream, Fucidin cream and Aron mix 9
Raman microscopy of crystalline structures in Diprosone cream, Fucidin cream and Aron mix 9 was performed using a Renishaw inVia Raman microscope (Renishaw Plc, UK), calibrated for peak position and intensity using a silicon reference block. Aron mix 9 was prepared 1 h in advance of analysis by weighing appropriate amounts of Fucidin cream, Diprosone cream and Diprobase cream in a 1:2:10 ratio into a glass dish and mixing thoroughly to ensure homogeneity. Samples of Diprosone cream alone, Fucidin cream alone and Aron mix 9 were mounted on Raman grade calcium fluoride slides for spectral analysis. Raman spectra were obtained using the x 100 long working distance magnification lens, a laser excitation wavelength of 785 nm, five accumulations per sample and an acquisition time of 10 s. Three replicate areas were scanned for each analysis and the single, most representative spectrum selected for presentation.

Quantitative analysis of betamethasone dipropionate and fusidic acid
Quantification of betamethasone dipropionate and fusidic acid was achieved using an Agilent 1260 Infinity quaternary pump coupled to an Agilent 1260 multi wavelength UV/Vis detector (Agilent Technologies, UK), set to 210 nm and 240 nm for fusidic acid and betamethasone dipropionate detection, respectively. Chromatographic analysis was performed using a reverse phase Kinetex TM C 18 column (5 µ particle size, 250 mm x 4.6 mm; Phenomenex, UK), a sample injection volume of 40 µL and a constant flow rate of 1 mL min -1 . A tertiary mobile phase system of sodium phosphate buffer (pH 3.1), HPLC grade acetonitrile and water (18.2 MΩ cm -1 ) was employed in the following ratios,

Human skin preparation
Excised human scrotal skin was obtained with informed consent from gender reassignment surgeries following ethical approval from the South London Research Ethics Committee (ethics No. 10/H0807/51). Skin samples were removed from storage (-20 °C) and left to thaw at ambient temperature, the subcutaneous fat was removed using a scalpel and samples were stored at -20 °C until required.

Ex vivo finite dose percutaneous absorption studies
Individually calibrated Franz cells (Soham Scientific, UK) with an average surface area of 1 cm 2 and average receiver volume of 3 mL Franz cells were assembled with human skin, the receiver chamber was filled with a mixture of PBS and ethanol (20 %) then skin samples were dosed with 10 μL of Diprosone cream, Fucidin cream or Aron mix 9 using a calibrated positive displacement pipette. Six replicates were performed for each formulation investigated (2 replicates per donor, 3 donors, matched across studies). To ensure contact with the membrane, the product was carefully spread over the surface of the skin with five clockwise, then anticlockwise, motions using the tip of a capillary piston. Aron mix 9 was prepared one hour in advance of dosing in the ratios of 1:2:10 of Fucidin cream, Diprosone cream and Diprobase cream, respectively using the method detailed in Section 2.2.1. Samples (200 μL) of the receiver fluid were taken at pre-determined intervals up to 24 h and replaced with fresh preheated receiver fluid. Drug quantification was achieved via HPLC UV.

Ex vivo finite dose skin distribution studies
Following the 24 h percutaneous absorption study, the residual formulation was collected from the donor chamber and skin surface by three sequential wipes with cotton buds (a dry cotton bud, a cotton bud soaked in acetonitrile then a final dry cotton bud) and two tape strips (Scotch Tape strips, 3 M Centre, USA) of the skin surface. The epidermal and dermal layers of skin samples were heat separated by placing the skin in an oven set to 60 °C for 1 minute before carefully peeling the epidermis and dermis apart (Kligman and Christophers, 1963). The skin layers were then placed in individual vials and the drug was extracted in aliquots of acetonitrile. Vials containing the samples were sonicated for 10 mins then placed on a roller mixer (Cole-Palmer, UK) for 18 h. To enable maximum recovery of fusidic acid and betamethasone dipropionate from all matrices, a second extraction was conducted for each sample. Drug quantification was achieved via HPLC UV.

Data treatment and statistical analysis
Drug concentration in the receiver fluid was corrected for previous sample removal and profiles constructed to present cumulative amount of drug permeated per unit area (μg cm -2 ) over the exposure period. Linear regression analysis was performed on infinite dose data sets to determined mean drug flux.
Experimental data were expressed as mean (n = 6) ± standard deviation (SD). Statistical analysis was performed using Prism 8.0 (GraphPad, USA). The Shapiro Wilk test was employed to determine the normality of all data sets. Non-parametric analysis for multiple comparisons was performed using Kruskal-Wallis and a Mann-Whitney test applied for post hoc analysis. Parametric analysis for multiple comparisons was performed using analysis of variance (ANOVA) and Tukey's post hoc test. Statistically significant differences were determined at a 95 % confidence interval (p ≤ 0.05).

Microscopic evaluation of the Aron mix and marketed preparations
Diprosone cream and Fucidin cream were analysed by Raman microscopy to understand whether drug was present in the formulations at saturated or sub saturated levels ( Figure 1). No potential drug particles were observed for Diprosone cream; thus it is likely that betamethasone dipropionate was formulated at submaximal drug thermodynamic activity in the formulation. Fusidic acid particles crystals, confirmed by Raman spectroscopy, were present in Fucidin cream in two distinct solid forms: rod shaped and square, planar shaped crystals ( Figure 2).. This indicated that Fucidin cream was formulated as a suspension at maximum stable thermodynamic activity in Fucidin cream. A model Aron mix formulation employed for the ex vivo skin permeation and penetration study was also investigated for the presence of drug crystals (Aron mix 9; 1:2:10). Fusidic acid drug crystals, confirmed by Raman microscopy, were evident in Aron mix 9 following a 13 fold dilution of Fucidin cream in Diprosone cream and Diprobase cream. No drug crystals were attributed to betamethasone dipropionate in Aron mix 9.

The impact of the degree of dilution of Diprosone cream or Fucidin cream in the Aron mix on drug transport across silicone membrane
The Aron formulations are tailored to suit the patient's age, weight and the severity of the condition, thus varying ratios of Fucidin cream and Diprosone cream diluted in a Diprobase cream base are extemporaneously prepared and dispensed. To investigate the impact on betamethasone dipropionate and fusidic acid transport across silicone membrane when Diprosone cream was diluted to varying extents, a series of formulations were prepared with a fixed concentration of Fucidin cream (4.35 % w/w) and varying concentrations of Diprosone cream (8.70-43.48 % w/w) in a Diprobase cream base (Aron mix 1-4). Silicone membrane is commonly used as a surrogate model membrane for skin to understand the effect of drug concentration or saturation on drug delivery from a topical formulation (Walters, 2002;Edwards et al., 2017). The effect of product dilution on drug thermodynamic activities in the Aron formulation, which influences drug delivery to the skin would be expected to be identified with silicone membrane, drug transport studies. Full details of the Aron mix compositions are presented in Table 1. The drug transport profiles for betamethasone dipropionate and fusidic acid from Diprosone cream, Fucidin cream and Aron mix 1-4 are presented in Figure 3. Comparatively, the same degree of dilution of Diprosone cream, but with a different base in Aron mix 8 (60.87 % Diprobase cream) resulted in a less proportional 7.1 fold decrease in Q26 compared to Diprosone cream alone (p < 0.05). This trend was likely to be attributable to the increasing proportion of Diprobase cream in the Aron mix, the base vehicle of Diprosone cream, as the proportion of Fucidin cream decreased, decreasing betamethasone dipropionate thermodynamic activity in the formulation to a greater extent than Fucidin cream.

Percutaneous absorption and skin retention of betamethasone dipropionate and fusidic acid from a model Aron mix
To confirm whether the observed effects of Aron mixes on the drug transport of betamethasone dipropionate and fusidic acid across silicone membrane resulted in significant differences in drug delivery to the skin, the percutaneous absorption and distribution of both drugs was evaluated in human skin following application of Aron mix 9, Fucidin cream and Diprosone cream. The use of human skin can provide understanding of additional formulation effects such as the impact of mixing the formulation on the performance of any chemical penetration enhancers present in the formulation (Walters, 2002). In this study human scrotal skin was used. The scrotum may be affected by atopic dermatitis and is histologically similar to skin from other body regions, although it is usually more permeable to drugs (Smith et al 1971). The tissue presents a barrier to drug penetration and although drug absorption many be relatively high, the insight gleaned from its use would be expected to be relevant for other body sites and thus it was selected as a suitable skin model for this study (Caserta et al 2019). Whilst Diprosone cream does not contain any notable potential penetration enhancers, Fucidin cream is formulated with a proportion of glycerol, a hygroscopic excipient which can increase the water holding capacity of the stratum corneum (Batt et al., 1988). Under conditions of low humidity, such as that created in stratum corneum affected by dry skin conditions, glycerol has been shown to interact with lipid bilayers to increase skin permeability of metronidazole (Björklund et al., 2013).
The recovery of betamethasone dipropionate and fusidic acid from the skin surface (residual formulation), epidermis, dermis and receiver fluid following the application of Diprosone cream alone, Fucidin cream alone and the Aron mix 9 is presented in Table 4. The absolute recovery of betamethasone dipropionate and fusidic acid ranged 94-108 % of the applied dose for all experiments conducted, falling within the OECD defined acceptable criteria (OECD, 2019). Betamethasone dipropionate was not detected in the receiver fluid following the application of Aron mix 9 and very low levels of the drug were recovered from the dermis. For clarity the total drug delivery (total drug content in the epidermis, dermis and receiver fluid) was used for statistical analysis as an indication of the change in total betamethasone dipropionate and fusidic acid delivery from Aron mix 9 when compared to Diprosone cream or Fucidin cream alone (Table 4).
Total betamethasone dipropionate delivery to the skin significantly decreased by 6 fold following the application of Aron mix 9 when compared to Diprosone cream alone (p < 0.05). This decrease was roughly proportional to the 6.5 fold dilution of Diprosone cream in the Aron mix. Comparatively, total fusidic acid delivery to the skin from Aron mix 9 significantly decreased by 5.4 fold compared to the application of Fucidin cream alone (p < 0.05), disproportionate to the 13 fold dilution of Fucidin cream in the Aron mix. This trend was consistent with the findings of the silicone membrane drug transport studies and suggests that whilst Diprobase cream was a suitable diluent for Diprosone cream to accomplish the intended reduction in betamethasone dipropionate absorption, the dilution effect on fusidic acid permeation was not predictable.
The percutaneous absorption profiles of betamethasone dipropionate and fusidic acid when Diprosone cream and Fucidin cream were applied alone or in Aron mix 9 are presented in Figure 6. Betamethasone dipropionate permeation was not detectable over 0-10 h, following the application of Diprosone cream alone to human skin. Following the application of Aron mix 9, betamethasone dipropionate was not detected in the receiver fluid over the entire experimental period. In comparison, fusidic acid was detectable in the receiver fluid at early timepoints following the application of Fucidin cream alone and its permeation profile was typical of an infinite dose study as would be expected from a suspension formulation, where the drug particles can dissolve in the formulation base to replace drug that had permeated the skin. This keeps the fusidic acid thermodynamic activity/saturation high in the formulation on the skin surface as drug is absorbed. Following the application of Aron mix 9, fusidic acid permeation was not detectable over 0-10 h, and considerably lower than that from Fucidin cream after 24 hours ( Figure 6). The low drug permeation rates from Aron mix 9 were consistent with the observed decrease in drug delivery to the epidermis and dermis. The premise for heavily diluting the TCS and antibiotic in the Aron mix is to allow an increase in the frequency of product application even to unaffected areas, up to six times daily (Aron, 2022b). Dilution (reduced drug concentration) of a topical formulation does not necessarily correlate with the extent of drug delivery to the skin and whilst the overall decrease in betamethasone dipropionate delivery to the skin was proportional to the degree of dilution of Diprosone cream, this was not the case for Fucidin cream. For Fucidin cream, the substantial dilution of the suspension formulation with a base dissimilar to the product resulted in disproportionate reductions fusidic acid delivery to the skin, making this difficult to predict in a clinical setting. The fusidic acid delivery from Aron mix 9 was not lowered to the same extent as the dilution factor. This is likely to be because of the fusidic acid particles present in Fucidin cream dissolving in the diluted Aron mix formulation raising the drug thermodynamic activity/saturation above that expected by a simple formulation dilution. However, the Aron mix 9 formulation still contained fusidic acid particles as shown by Raman microscopy (Section 3.1). In theory, higher fusidic acid delivery from Aron mix 9 might therefore be expected, perhaps matching that of the Fucidin cream suspension. It is possible that the dissolution rate of the fusidic acid crystals in the Aron mix 9 formulation may not have been rapid enough to fully maintain the high drug thermodynamic activity in the formulation. The fusidic acid crystals observed in Aron mix 9 could potentially dissolve further in the Aron mix base with time, altering the formulation performance. Consideration of product stability following the preparation of Aron regime formulations was outside the scope of this study. Simultaneous to the decreasing fusidic acid thermodynamic activity in Aron mix 9, mixing Fucidin cream with Diprobase cream and Diprosone cream is likely to have resulted in a decrease in the thermodynamic activity of glycerol in the extemporaneous mix, compared to Fucidin cream alone which may have affected its delivery to the skin. Thus, the extent to which total fusidic acid delivery to the skin was altered is likely to have been influenced by (i) the decrease in drug thermodynamic activity and (ii) the decrease in thermodynamic activity of potential penetration enhancers in Aron mix 9.
S. aureus infections are typically localised to the skin surface and stratum corneum (Arikawa et al., 2002). However, skin colonisation of S. aureus has been found to extend beyond the epidermal barrier and into the dermis in lesional sites of atopic dermatitis patients (Nakatsuji et al., 2016). Thus, successful treatment may require delivery of fusidic acid in sufficient concentrations into the skin to decolonise the affected sites and prevent recurrent skin infections. The summary of product characteristics for Fucidin cream advises that fusidic acid concentrations of 0.03-0.12 µg mL -1 (equivalent to 58-232 nM) are sufficient to inhibit nearly all strains of S. aureus (Leo Laboratories Ltd, 2019). The recovered drug levels in the epidermis and dermis are presented in Table 5 for comparison with the reported MIC for fusidic acid.
The drug concentration in the epidermis was 3,365 fold and 1,221 fold greater than the MIC for fusidic acid, following the application of Fucidin cream alone and Aron mix 9, respectively. Drug concentration in the dermis was 528 fold and 236 fold greater than the MIC of fusidic acid, following the application of Fucidin cream alone and Aron mix 9, respectively. Care should be taken however before assuming that the drug was delivered to the epidermal and dermal tissue in sufficient concentrations, following a single application of Aron mix 9, for activity against S. aureus, as drug binding to keratin or other proteins may affect its antibiotic efficacy. It is also important to consider this finding in the context of the Aron regimen, which involves repeat applications of the product to affected and unaffected areas of the skin up to six times a day during the initial phase of treatment (minimum of two weeks). Clinical recommendations widely caution against the extended use of topical antibiotics in the treatment of clinically infected atopic dermatitis (Bath-Hextall et al., 2010;Eichenfield et al., 2014;Wollenberg et al., 2018) and where offered, treatment should be applied to localised areas only for a maximum of seven days (NICE, 2021). To fully evaluate the role of the Aron regimen in the treatment of severe cases of infected (or uninfected) atopic dermatitis, controlled investigations into the clinical efficacy of combined TCS and antibiotic therapies are required.

Conclusions
Compounding a complex dual therapy of a TCS and a topical antibiotic in an emollient base reduced the expected drug transport profiles, percutaneous absorption and skin retention of betamethasone dipropionate and fusidic acid relative to the individual marketed products. For Diprosone cream, the decrease in betamethasone dipropionate transport across silicone membrane and delivery to the skin largely correlated to the degree of dilution of the product in the Aron mix. For Fucidin cream, the impact of dilution did not correlate with the extent of reduction in fusidic acid transport across silicone membrane or skin absorption and was attributed to several formulation effects occurring in the complex extemporaneous mixtures. These effects are difficult to predict in a clinical setting. Preliminary investigations revealed that fusidic acid was delivered to the epidermal and dermal tissue in relatively high concentrations following a single application of a model Aron mix. To evaluate whether these findings correlate with clinical efficacy, further studies are required. In addition, the effect of frequent daily applications of the Aron mix to widespread areas on TCS side effects and antimicrobial resistance should be investigated.   Data are shown as the mean of six replicates (± SD).  Aron mix 9 (□ denotes BDP and ○ denotes FA). Aron mix 9 contained Fucidin cream, Diprosone cream and Diprobase cream mixed in a 1:2:10 ratio. Data are shown as mean ± SD (n = 6).   Table 4: The distribution of betamethasone dipropionate (BDP) and fusidic acid (FA) in the residual formulation, skin layers, receiver fluid and total absorbed (sum of epidermis, dermis and receiver fluid)

Figures and Tables
following the application of Diprosone cream alone, Fucidin cream alone or Aron mix 9. Data are shown as mean ± SD (n = 6). * denotes a significant difference when total BDP or FA delivery from Aron mix 9 was compared to total drug delivery from Diprosone cream or Fucidin cream, respectively.