2.1.

Investigational Ointment

The investigational cosmetic ointment provided by Klosterfrau Healthcare Group, Cologne, Germany, contains a hypericum perforatum extract (1.5% w/w) that is rich in hyperforin. The extract also includes the following ingredients (International Nomenclature of Cosmetic Ingredients): aqua, petrolatum, propylene gycol, caprylic/capric triglyceride, PEG-20 glyceryl stearate, cetyl alcohol, glyceryl stearate, panthenol, phenoxyethanol, tocopherol, tocopheryl acetate, and allantoin. On average, $29.2\pm 14.5\mathrm{g}$ of the ointment were applied by the volunteers on their forearms.

2.2.

Subjects

Fourteen volunteers with dry or atopic skin were investigated before and after the application of the cream. Seven subjects were recruited at the Department of Dermatology of the Charité-Universitätsmedizin Berlin and seven subjects at the Department of Dermatology of University Medical Center Freiburg. The treatment period was 14 days in Freiburg and 28 days, with an intermediate visit on day 14, at the Charité.

The study design was evaluated and approved by the Ethics Committees of the Charité-Universitätsmedizin Berlin and the University Medical Center Freiburg. All volunteers had given their written informed consent for participation in the study.

2.3.

Study Procedure

For each volunteer, a mask of the inner forearm was prepared on an overhead foil using natural markers (i.e., naevi) to later find the same areas at all measurement time points.

First, the evaluation of the subjective parameters was performed and the transepidermal water loss (TEWL) and SC moisture were measured. Subsequently, the LSM investigations were performed. The SC lipids were extracted in proximity to the investigated area to avoid contaminations with marker and immersion oil in the extraction fluid.

2.4.

Parameters of the LSM Investigations

For laser scanning microscopy, seven volunteers were investigated at day zero and after 28 days using a commercially available LSM system (VivaScope® 1500 Multilaser, MAVIG, Munich, Germany) with the reflectance and fluorescent modes. The system is equipped with three low-power diode lasers emitting at 488, 658, and 785 nm (max. 15 mW, laser class 1M) and a lens with a numerical aperture of 0.9. In this study, the wavelength of 488 nm was used for the investigation in the fluorescence mode and 658 nm for the reflectance confocal microscopy (RCM). The device provides a lateral resolution of 0.5 to $1.0\mu \mathrm{m}$ and an axial resolution of 3 to $5\mu \mathrm{m}$ with a maximum imaging depth of 200 to $250\mu \mathrm{m}$. A drop of fluorescein was applied, and after a 5-min penetration time, it was removed to enable the fluorescent measurements. A drop of immersion oil was applied to the surface of the skin in order to minimize spherical aberration. Then, an objective steel ring with an adhesive plastic ring was applied to the skin. First, a dermatoscopic image was obtained by using an integrated dermatoscopic camera. Subsequently, ultrasound gel was applied between the objective ring and the objective, and the LSM imaging process was started. Horizontal composite images of $8\times 8\mathrm{mm}$ (confocal mosaics) were obtained on the surface in the fluorescence and reflectance modes to present the SC structure. Four representative images of the SC were taken in the fluorescent mode. Furthermore, sequential horizontal images of $500\times 500\mu {\mathrm{m}}^2$ were obtained in reflectance mode on the areas of special interest, starting at the SC and continuing through the stratum granulosum (SG), stratum spinosum (SS), and the dermal-epidermal junction (DEJ) to the superficial dermis. The frame rate was nine frames per second.

From the mosaic of the surface, the uniform arrangement of the furrows was determined. Zero was related to a normal uniform arrangement, 1 to slight derangement, and 2 to strong derangement.

The skin surface dryness was determined using the fluorescent images of the SC. A score from 1 to 3 was developed; 1 = well-tended skin, 2 = slightly dry, 3 = strongly dry and disturbed corneocyte structure.

The quality of the honeycomb structure of the SS was determined by a score, taking the first images of the SS; 0 = regular structure, 1 = slightly irregular, 2 = strongly irregular.

The spongiosis is mainly an intercellular edema that is characterized by dark areas; 1 = no spongiosis, 2 = small spongiosis, 3 = strong spongiosis.

Other criteria for inflamed skin are the visibility of the epidermal papillae or the elongated papilla. As most of the papillae were hardly visible, the size of the papillae was not estimated. The score used for the recognition of the papilla was 1 = clear visibility, 2 = difficult to recognize, 3 = no demarcation.

2.5.

Subjective Parameters

The subjective parameters, like skin dryness, itching, and tension, were determined at the time points T0 ($n=14$), T14 ($n=14$), and T28 ($n=7$); the volunteers were asked to estimate the parameters on a scale from 0 to 10, whereby 0 was related to no dryness, no itching, and no tension and 10 to the strongest effects of the parameters.

2.6.

Biophysical Skin Parameters

To characterize the skin barrier function, the TEWL (Tewameter TM 300, Courage + Khazaka electronic GmbH, Cologne, Germany) was measured at the time points T0 ($n=14$), T14 ($n=14$), and T28 ($n=7$). The corneocyte moisture was measured using the corneometer CM 825 (Courage + Khazaka electronic GmbH, Cologne, Germany) at the same time points.

2.7.

Skin Lipid Determination

It is well known that skin lipids are strongly modified in the skin of atopic dermatitis patients.¹⁷ To determine the skin lipid profile, lipids were directly extracted from the upper skin using ethanol as described previously.^18,19 Skin lipid analysis was performed at T0 and T28 in seven volunteers. From each skin area (diameter: $\sim 1.77{\mathrm{cm}}^2$), the skin lipids were extracted. The following procedure was repeated in triplicate in order to obtain 9 mL of skin lipid extract. A centrifugation tube filled with 3 mL solvent was placed on the skin and the extraction was performed by shaking the arm with the tube for 1 min. This was repeated three times on the same measuring area. Extracts were dried under a continuous stream of nitrogen and reconstituted in $300\mu \mathrm{L}$ n-hexane/methanol, $2:3$ (v:v). Samples were stored at $-20°\mathrm{C}$ until the analysis with high-performance thin-layer chromatography (HPTLC) was conducted.

The chromatographic analysis of the lipid profile was performed according to the procedures of Farwanah et al. as well as Schellenberg and Kabrodt^20,21 with slight modifications. For the analysis of the skin lipid profile, $10\mu \mathrm{L}$ of the skin extracts and the standard solutions were applied as 6 mm bands under a flow of nitrogen on silica gel plates (HPTLC, Silica Gel 60, Merck, Darmstadt, Germany, $10\times 20\mathrm{cm}$, prewashed with methanol/chloroform (v:v, $1:2$) using an HPTLC autosampler (ATS4, CAMAG, Muttenz, Switzerland). For the analysis of the cream lipid profile, the cream was dissolved in n-hexane/methanol, $2:3$ (v:v). Chromatographic separation was completed with a nine-step gradient using an automated development chamber (AMD 2, CAMAG). Elution solvents were methanol (Roth, Karlsruhe, Germany), chloroform (Grüssing, Filsum, Germany), toluene (Th. Geyer, Renningen, Germany), and n-hexane (Merck). 4 M acetic acid (Roth) was used as a preconditioning agent to improve the band sharpness of oleic acid. Lipids were quantified after staining with 10% ${\mathrm{CuSO}}_4$ (Roth), 8% ${\mathrm{H}}_3{\mathrm{PO}}_4$ (Merck), aqueous solution, and charring at 180°C for exactly 8 min using a TLC Wavelength- Scanner 3 (CAMAG) equipped with WinCATS software (version 1.4.4.6337) at a wavelength of 600 nm. For this purpose, L-$\alpha$-phosphatidylcholine (PC), cholesterol (Chol), oleic acid, glyceryl trioleate, cholesteryl oleate, sodium cholesteryl sulfate, squalene (SQ), sphingomyeline (SM) (all Sigma Aldrich), ceramide [NS] (Sederma, Nettetal, Germany), ceramide [NP] (Cer [NP]), and ceramide [AP] (Cer [AP]) (Evonik, Essen, Germany) were used. All standards were dissolved in methanol/chloroform (v:v, $1:1$). For the determination of single compounds in SC extracts, the standard ceramides Cer [NP] and Cer [AP] were used for the quantification of two substances, each described as Cer [NP1] and Cer [NP2], and Cer [AP1] and Cer [AP2]. The amounts of L-$\alpha$-phosphatidylcholine and sphingomyeline were determined as a sum (sum PC+SM). Calibration was conducted three times using 10 calibration solutions of the standard mixture.²²

2.8.

Data Analysis and Statistical Evaluation

The statistical results were calculated using Microsoft® Office Excel for determining the mean values, balances, and standard errors as well as SPSS® 19 for Windows. The descriptive exploratory data analysis of the related variables was subjected to the Friedman test and the Wilcoxon test. A significant difference was found in the case of $p\le 0.05$.

3.1.

Laser Scanning Microscopy

The SC structure representing the dryness was determined using the fluorescent images of the SC, which were measured at 488 nm (Fig. 1). A honeycomb pattern of the corneocytes, as shown in Fig. 1(b), represents well-tended skin. Most of the volunteers started with a dry and disturbed SC structure, visible by highly reflecting corneocytes, which are desquamated cells [Fig. 1(a)]. Partial depressions of the surfaces were also visible at the beginning of the study.

Fig. 1

Selected laser scanning microscopy (LSM) images of the stratum corneum (SC) of two volunteers in the fluorescence mode. An example of (a) a dry and disturbed corneocyte structure and (b) an almost well-tended skin.

Using the reflectance mode at 658 nm, mosaics of the skin at the depth of the SS could be obtained (Fig. 2). The healthy skin showed a uniform arrangement of a polygon mesh of the furrows [Fig. 2(a)], which disappeared in atopic skin with moderate lesions [Fig. 2(b)]. In parts of the images of the lesions, disrupted skin was visible.

Fig. 2

Mosaics of the skin surface making the furrows on the surface visible. (a) An almost uniform arrangement of a polygon mesh of the furrows is visible. (b) Strongly atopic skin lost the uniform arrangement and mainly parallel furrows are visible.

To obtain semiquantitative data, a score was developed for the skin surface dryness and the arrangement of furrows. The results of the evaluation are shown in Fig. 3.

Fig. 3

(a) score of SC dryness after analysis of the LSM images in the fluorescence mode: 1 corresponds to a well-tended skin, 2 to slightly dry, and 3 to strongly dry and disturbed corneocyte structure. (b) score of uniform arrangement of the furrows after analysis of the LSM images in the reflectance mode: 0 corresponds to a normal uniform arrangement, 1 to a slight derangement, and 2 to a strong derangement ($n=7$).

For most of the volunteers, a trend of reduction in the dryness was obtained from 2.4 to 2.0 after 28 days of application (Fig. 3, left). The uniform arrangement of the furrows did not change over the investigated time period (Fig. 3, right).

After taking the mosaic images, z-scans in the reflectance mode were performed to investigate the epidermis until the DEJ zone. In Fig. 4, images of a healthy but dry skin and atopic skin with a moderate erythema are shown.

Fig. 4

Reflectance confocal microscopy (RCM) measurement of dry (upper panel) and moderately inflamed atopic skin (lower panel).

Healthy skin starts with an intact SC and SG with regularly formed cells in the SG. The atopic skin at this depth showed a disrupted skin barrier where parts of the cells of the SG lost their regular form. In healthy skin, the honeycomb structure in the SG and SS was visible, whereas in the atopic skin, spongiosis in terms of dark areas relative to the surrounding epithelium appeared. The intercellular spaces between the keratinocytes were larger than normal.

In atopic skin, the SS was extended and, thus, the epidermis thickness was increased. In healthy skin, the top of the basal cell layer at $15\mu \mathrm{m}$ was characterized by pigmented skin cells, which appeared brighter and surrounded the blood capillaries (arrow). The pigmented cells provided a higher reflective index and, therefore, appeared brighter. They were absent in inflamed atopic skin.

Furthermore, the papillae consisting of basal cells were visible in intact skin at the DEJ zone. These may not be visible in inflamed atopic skin. The blood vessels appeared directly and were partially elongated, and a high throughput of erythrocytes could be observed during the z-scan measurement. The images seemed to be blurred. This continues in the dermis (D) down to a depth of $150\mu \mathrm{m}$, where elastic fibers and collagen should be visible as presented in intact skin at $90\mu \mathrm{m}$. Because of the missing structure of the papillae and the blurry images of the DEJ zone, the diameter of the papillae and epidermis thickness could not be estimated for inflamed atopic skin.

To present this loss in structure at the DEJ in more detail, images before and after treatment of one volunteer starting with a superficial moderate eczema are presented in Fig. 5. The eczema was reduced after two weeks of treatment, and this was reflected clearly in the RCM images.

Fig. 5

RCM images of stratum papillae at T0 (a) and T28 (b) of one patient. The supposed blurring of the image is clearly visible in the left image. The bright papillae rims are almost lost and the cellular borders have vanished. The left image is taken at $84\mu \mathrm{m}$ and the right one at $65\mu \mathrm{m}$, indicating the epidermal thickening.

To get some statistical data for the honeycomb pattern and spongiosis of the SS, scores were developed. The mean values of these scores are shown in Fig. 6.

Fig. 6

Spongiosis and honeycomb structure of the stratum spinosum imaged by RCM of the inner forearm before and after the application of the ointment. Score spongiosis: 1 = no spongiosis, 2 = weak spongiosis, 3 = strong spongiosis. Score honeycomb structure: 0 = regular structure, 1 = slightly irregular, 2 = strongly irregular ($n=7$).

Most of the volunteers started with a relatively high irregular structure and a moderate spongiosis. After application of the investigational ointment, the spongiosis was significantly reduced and the quality of the honeycomb pattern was also improved (not significant).

Going deeper into the skin, the bright papillae rimed at the DEJ in intact skin. In atopic skin, the bright papillary could be absent. The score values at the beginning at T0 were $\sim 2.2$, which indicates that they are difficult to recognize and partly that there was no demarcation. At T28, the score improved to $\sim 1.8$, which means that the papillae were still difficult to recognize but partly clearly visible. The changes were not significant.

3.2.

Biophysical Skin Parameters

The skin parameters, TEWL and corneocytes moisture, relative to the initial values are shown in Fig. 7. The corneocyte moisture significantly increased between T0 and T14. From T14 to T28, the values were enhanced slightly (not significant).

Fig. 7

SC moisture and transepidermal water loss relative to the initial values ($\text{mean}\pm \mathrm{SE}$) at three visits (*$p<0.05$) [$n(\mathrm{T}0,\mathrm{T}14)=14$, $n(\mathrm{T}28)=7$].

The TEWL values showed no significant changes over the study time. The values showed high variations, which was, inter alia, due to the genesis of new erythema at T28.

The 15 SC lipids were analyzed before (T0) and after the application of the hyperforin-containing ointment at 28 days (T28). Several lipids increased after the application of the ointment, but only for Cer [NS], a trend ($p<0.1$) and for Chol, could a significant increase be observed. In Fig. 8, the lipid concentrations relative to the initial values (V2/V1) are shown.

Fig. 8

SC lipids at T28 relative to the initial values at T0 are shown. *$p<0.5$. (*) $p<0.1$ ($n=7$).

3.3.

Subjective Skin Parameters

In Fig. 9, the subjective parameters are shown. All parameters decreased significantly from T0 to T14 and improved further to T28, but not significantly.

Fig. 9

$\text{Mean values}\pm \mathrm{SE}$ of the subjective skin parameters, where 0 is related to no dryness, no tenderness, and no itching and 10 reflects a strong effect of the parameters ($n=7$).