1.

Introduction

Glucocorticoids (GC) are frequently used for the topical therapy of inflammatory skin diseases. The most common adverse effect of topical corticosteroids is cutaneous atrophy, which is characterized by a decrease in skin thickness and loss of elasticity causing cutaneous transparency, increased fragility, and telangiectasia. Topical GC treatment is associated with an increased permeability and transepidermal water loss¹ that indicates a disturbed skin barrier function.^2,3 Histopathological findings include flattening of the dermoepidermal junctions^4,5 and a reduction of epidermal thickness and number of fibroblasts.^1,6,7 Transmission electron microscopic measurements showed a loss of ground substance after 6 weeks of occlusive clobetasol application leading to decreased spaces between collagen and elastic fibers resulting in a more compact papillary and reticular dermis, whereas collagen fibers appeared less woven but rather regularly arranged.⁸

Antioxidants neutralize free radicals and, therefore, protect the skin from free radical-induced oxidative stress. The equilibrium of both antioxidants and free radicals forms the antioxidant status of the skin. Using carotenoids as marker substance of the antioxidant status in the human skin changes in the antioxidant status of the skin due to free radical formation induced by UV exposure or further internal and external factors were assessed noninvasively as described previously.^9,10 Previous data indicate that GC decrease the activity of glutathione peroxidase and enhance the accumulation of reactive oxygen species in hippocampal neurons,^11,12 whereas the effect of topical GC treatment to the antioxidant status of the skin has remained unclear.

Several noninvasive methods for assessment of aspects of skin atrophy have been described. Established imaging techniques, such as high-frequency ultrasound (US) and optical coherence tomography (OCT), are useful for evaluating dermal and epidermal thinning, respectively. Using OCT, it was shown that after 3 days of topical steroid treatment a significant reduction in epidermal thickness can be found.^1,13 However, these techniques cannot deliver the subcellular resolution that is required to evaluate important features such as dyskeratosis, morphological changes of the epidermal cell layers, or changes in the dermal fibrillary network. Confocal laser scanning microscopy (CLSM) and multiphoton tomography (MPT) are in vivo imaging techniques with very high spatial resolution that allow assessment of cell morphology. MPT is based on the nonlinear excitation of endogenous fluorophores, among others NAD(P)H, flavoproteins, keratin, lipofuscin, elastin, collagen, melanin, and porphyrins. Due to their noncentrosymmetric structure, collagen fibers generate a strong second-harmonic generation (SHG) signal and can be imaged within the dermis with SHG microscopy.¹⁴ Recently, SHG was also used for the investigation of collagen-fiber orientation and their structural changes in fibrotic collagen,¹⁵ human dermis,^16–19 and keloid tissue.²⁰ The elasticity and tensile strength of the dermis is dependent on a fibrillar network consisting of mainly collagen and elastin, but also proteoglycans and glycoproteins. Some of the clinical features of GC-induced skin atrophy have been ascribed to the loss of dermal collagen and elastin.^21,22 The fiber network density and morphology have been used to access age-related alterations found in the dermis, which are particularly related to alterations in the collagen network. Upon aging, the collagen fibers thicken while the elastin network shows fewer microfilaments and a loosening of the fibers.^5,23 The second-harmonic-to-autofluorescence (SHG to AF) aging index of dermis (SAAID) is defined as (SHG − AF) / (SGH + AF) and has previously been described as a noninvasive way to measure the aging processes in the human skin in vivo.²⁴ It is based on the simultaneous determination of AF and SHG signal intensities and indicates the fibrous status of the dermis. Previous investigations showed that the SAAID decreases with age²⁴ and after photoaging processes in the sun-exposed skin areas.²⁵

The aim of this study was to examine the effects of topical GCs on stratum corneum, epidermal, and dermal structures in vivo using high-resolution imaging techniques. In addition to that the antioxidant status of the epidermis was assessed as an indicator for free radical formation induced by GC treatment.

2.

Materials and Methods

2.1.

Study Design

This phase I, single-center, explorative, randomized, investigator-blinded, negative-controlled trial with intraindividual comparison was conducted at the Charité-Universitätsmedizin Berlin, Department of Dermatology and Allergy. Eligible subjects were recruited from January 29, 2015 to February 27, 2015. The last subject finished the trial on April 28, 2015.

2.1.1.

Randomization and blinding

Randomized assignment to two different measurement groups and treatment assignment to test sites was conducted prior to the first measurements. The four test sites A, B, C, and D (group 1: $28{\mathrm{cm}}^2$, each; group 2: $20{\mathrm{cm}}^2$, each) were marked on the lower volar surface of the arms, A and B on the right arm and C and D on the left arm.

The subjects were exposed once daily over a 28-day treatment period including weekends to each of the following treatments with $\sim 1.7\mathrm{mg}/\mathrm{c}{\mathrm{m}}^2$ on the marked test sites: clobetasol propionate 0.05% (CP), betamethasone dipropionate 0.064% (BDP), and petrolatum ointment, which was used as the vehicle of the steroid formulations, as negative control. One test site remained untreated. The steroid treatment with CP and BDP was conducted over a 28-day period and followed by a 7-day treatment-free period starting at day 29 up to day 36. For allocation of the groups and treatments, a computer-generated list was used. Randomization was performed successively using sealed randomization envelopes. Investigators involved in the study assessments were blinded to the identity and allocation of the treatments.

2.1.2.

Subjects

A total of 16 healthy male subjects of skin types I to III, aged between 20 and 40 years, entered treatment phase for the trial (Fig. 1). Informed consent was obtained from all subjects. Main exclusion criteria included clinical skin atrophy, telangiectasia or striae on volar arms, presence of any skin condition or coloring, e.g., marked suntan, hyperpigmentation, scars, tattoos, or body art, that would have interfered with test sites or the response or assessment. Further exclusion criteria were history or current evidence of infection, eczema, or other relevant skin diseases (e.g., atopic dermatitis or known contact allergy), significant history or current evidence of chronic infectious disease, system disorder (especially hypertension or circulatory disease), or organ dysfunction and known predisposition for hypertrophic scar formation.

Fig. 1

Flow diagram of participants.

2.1.3.

Ethical approval

Prior to initiation of the study, the protocol and the subject informed consent form were approved by the independent Ethics Committee of the State Office of Health and Social Affairs Berlin (LAGeSo). The study was registered at the European Union Drug Regulating Authorities Clinical Trials Database (EudraCT 2014-001450-42). The study was conducted according to the principles of the Declaration of Helsinki (1996) and Good Clinical Practice Guidelines.

2.1.4.

Assessments

Treatments were applied, according to the randomization plan, once daily by study nurses on the respective test areas. The study products were administered by a 5-s lasting gently rubbing application independently and blinded from investigators.

Measurements by US, OCT, LSM, MPT, and resonance Raman spectroscopy (RRS) were performed on the marked areas A, B, C, and D at baseline, day 29 [at end of treatment (eoT)], and day 36 (7 days after eoT). The obtained measuring parameters included stratum corneum thickness, epidermal thickness, full skin thickness (defined as the sum of epidermal and dermal thickness), keratinocyte size and density, nucleus-to-cytoplasm ratio, skin surface classification, cell nucleus visibility on the skin surface, SAAID, and cutaneous carotenoid concentration. All observers were blinded to the treatment group for the analysis of obtained images and spectra.

3.

Results

3.1.

Skin Thickness

Overall, a reduction of both epidermal and full skin thickness was observed in GC- and vehicle-treated areas as well as in untreated skin ($p<0.001$) (Figs. 2 and 3 and Table 3). GC treatment induced an enhanced reduction in epidermal and full skin thickness compared to vehicle-treated and -untreated skin areas, whereas CP induced a slightly higher thinning of epidermal and full skin thickness compared to BDP after 28 days of treatment. Stratum corneum thickness was significantly reduced in GC and vehicle-treated sites, whereas only a slight thinning effect was observed in untreated skin.

Fig. 2

Effect of GC treatment on stratum corneum, epidermal, and full skin thickness assessed by OCT, CSLM, and US. Day 1 is at baseline, day 29 is the last day of treatment, and day 36 is a 1-week follow-up.

Fig. 3

(a) Exemplary vertical images of epidermal thickness measured by OCT at baseline, (b) after 28 days of treatment at day 29, and (c) 7 days after eoT in a CP-treated skin area with arrows marking the dermoepidermal junction. A notable thinning of the epidermis can be observed after 28 days of CP treatment (b) with a partial recovery 7 days after eoT (c).

Table 3

Mean value and standard deviation (SD) of changes (μm) from baseline to day 29 of full skin thickness by US, of epidermal thickness by OCT, and of stratum corneum thickness by LSM at baseline, day 29, and day 36. P-values describe the changes from baseline to day 29 after treatment of the skin with CP, BD, and petrolatum as well as in untreated skin areas.

	CP 0.05%	BDP 0.064%	Petrolatum ointment	Nontreated
Change from baseline in full skin thickness ($\mu \mathrm{m}$)
Day 29
$N$	16	16	16	16
$\text{Mean}\pm \mathrm{SD}$	$-277.8\pm 151.2$	$-276.4\pm 85.6$	$-149.8\pm 91.5$	$-105.1\pm 101.3$
Median	$-316.0$	$-267.0$	$-144.5$	$-66.5$
$P$-value	$<0.001$**	$<0.001$**	$<0.001$**	$<0.001$**
Day 36
$\text{Mean}\pm \mathrm{SD}$	$-173.7\pm 126.4$	$-158.3\pm 101.3$	$-121.4\pm 104.8$	$-50.3\pm 101.4$
Median	$-166.5$	$-166.0$	$-98.0$	$-51.5$
Change from baseline in epidermal thickness ($\mu \mathrm{m}$)
Day 29
$N$	16	16	16	16
$\text{Mean}\pm \mathrm{SD}$	$-41.6\pm 14.7$	$-38.6\pm 13.6$	$-16.5\pm 11.3$	$-12.8\pm 14.3$
Median	$-41.0$	$-40.0$	$-20.0$	$-10.0$
$P$-value	$<0.001$**	$<0.001$**	$<0.001$**	$<0.007$**
Day 36
$\text{Mean}\pm \mathrm{SD}$	$-22.3\pm 16.9$	$-19.9\pm 12.8$	$-11.9\pm 10.0$	$-5.8\pm 9.1$
Median	$-18.0$	$-20.0$	$-10.0$	$-4.0$
Change from baseline in stratum corneum thickness ($\mu \mathrm{m}$)
Day 29
$N$	8	8	8	8
$\text{Mean}\pm \mathrm{SD}$	$-2.608\pm 1.013$	$-2.333\pm 2.317$	$-2.533\pm 2.643$	$-0.2\pm 1.014$
Median	$-2.7$	$-3.0$	$-3.0$	0
$P$-value	0.009*	0.028*	0.003*	0.804
Day 36
$\text{Mean}\pm \mathrm{SD}$	$-2.167\pm 1.976$	$-2.779\pm 1.923$	$-2.283\pm 2.239$	$-0.829\pm 1.056$
Median	$-2.5$	$-2$	$-2.283$	$-0.75$

^*Significance at p<0.05.

^**Significance at p<0.001.

All treatment areas showed a partial recovery 1 week after eoT at day 36, but none of them showed a complete recovery to baseline thickness.

3.2.

Effect of GC Treatment on Epidermal Cell Density and Morphology

Analysis of the MPT-AF images showed significant changes of epidermal cell number and size in GC-treated skin areas, as well as petrolatum-treated and nontreated areas, respectively (Table 4, Figs. 4Fig. 5–6). CP and BDP treatment induced a notably higher increase in cell density in treated areas compared to petrolatum-treated or nontreated areas, whereas CP showed higher structural effects than BDP. The increase in cell density was corresponding to a significantly reduced size of keratinocytes of the stratum granulosum in both GC-treated sites. The nucleus-to-cytoplasm ratio showed an overall increase in all skin sites at day 29 with the highest significant increase in GC-treated skin. All morphologic changes show a partial recovery 7 days after eoT.

Table 4

Mean values and SD of cell density, keratinocyte size, and nucleus-to-cytoplasm ratio in stratum granulosum and stratum spinosum per 100×100  μm2 measured by MPT-AF at baseline, day 29 (at the eoT), and day 36 (7 days after eoT).

	CP	BDP	Petrolatum ointment	Nontreated
N	8	8	8	8
Cell count stratum granulosum per $100\times 100{\mu \mathrm{m}}^2$a
Baseline	$23.94\pm 2.51$	$24.56\pm 3.03$	$23.69\pm 2.39$	$24.69\pm 2.51$
Day 29d	$32.38\pm 2.94$	$32.88\pm 3.69$	$26.63\pm 2.43$	$27.38\pm 2.86$
Day 36e	$29.56\pm 2.90$	$27.94\pm 4.14$	$25.94\pm 3.79$	$26.44\pm 2.82$
$P$-value (change from baseline to day 29)	$<0.001$**	$<0.001$**	0.0223*	0.0357*
Cell count stratum spinosum per $100\times 100{\mu \mathrm{m}}^2$a
Baseline	$32.69\pm 5.26$	$31.50\pm 3.28$	$31.75\pm 4.50$	$33.19\pm 2.67$
Day 29d	$39.38\pm 4.70$	$38.44\pm 5.07$	$34.50\pm 4.60$	$35.31\pm 4.57$
Day 36e	$35.75\pm 3.17$	$35.31\pm 3.85$	$32.75\pm 3.64$	$34.88\pm 5.03$
$P$-value (change from baseline to day 29)	0.001**	$<0.001$**	0.0908	0.1877
Keratinocyte size stratum granulosum ($\mu \mathrm{m}$)b
Baseline	$20.33\pm 0.77$	$19.66\pm 1.13$	$19.97\pm 0.98$	$20.06\pm 1.72$
Day 29d	$18.00\pm 0.66$	$18.32\pm 1.05$	$19.22\pm 0.91$	$19.17\pm 0.69$
Day 36e	$18.61\pm 0.67$	$18.36\pm 0.62$	$19.28\pm 0.86$	$18.97\pm 0.40$
$P$-value (change from baseline to day 29)	$<0.001$**	0.005*	0.0992	0.0165*
Keratinocyte size stratum spinosum ($\mu \mathrm{m}$)b
Baseline	$12.21\pm 0.80$	$11.87\pm 0.99$	$11.95\pm 0.90$	$12.50\pm 0.31$
Day 29d	$11.08\pm 0.45$	$11.79\pm 1.34$	$11.70\pm 0.99$	$11.14\pm 1.27$
Day 36e	$11.44\pm 0.97$	$11.20\pm 1.07$	$11.31\pm 1.20$	$11.62\pm 1.12$
$P$-value (change from baseline to day 29)	0.017*	0.867	0.587	0.005*
Nucleus-to-cytoplasm ratio stratum granulosumc
Baseline	$0.33\pm 0.04$	$0.33\pm 0.04$	$0.33\pm 0.05$	$0.33\pm 0.04$
Day 29d	$0.42\pm 0.10$	$0.33\pm 0.05$	$0.34\pm 0.05$	$0.36\pm 0.04$
Day 36e	$0.38\pm 0.04$	$0.37\pm 0.04$	$0.34\pm 0.04$	$0.34\pm 0.04$
$P$-value (change from baseline to day 29)	$<0.001$**	$<0.001$**	0.064	0.015*
Nucleus-to-cytoplasm ratio stratum spinosumc
Baseline	$0.43\pm 0.64$	$0.47\pm 0.07$	$0.44\pm 0.05$	$0.47\pm 0.07$
Day 29d	$0.53\pm 0.07$	$0.48\pm 0.08$	$0.48\pm 0.08$	$0.51\pm 0.08$
Day 36e	$0.47\pm 0.08$	$0.50\pm 0.07$	$0.49\pm 0.07$	$0.44\pm 0.07$
$P$-value (change from baseline to day 29)	$<0.001$**	0.414	0.009*	0.088

^aCell count of the stratum granulosum and spinosum per 100×100μm2 determined in twofold measurements using multiphoton autofluorescence excitation microscopy by multiphoton tomography (MPT-AF).

^bKeratinocyte size stratum granulosum (μm) determined in sixfold measurements using multiphoton autofluorescence excitation microscopy by multiphoton tomography (MPT-AF).

^cPlasma-to-nucleus ratio stratum spinosum determined in sixfold measurements using MPT-AF.

^dDay 29 was the last day of treatment.

^eDay 36 assessment was done 1 week after last treatment.

^*Significance at p<0.05.

^**Significance at p<0.001.

Fig. 4

Keratinocyte size in the stratum granulosum and stratum spinosum of the four treated sites at baseline, day 29 and day 36. The highest reduction in size was observed in clobetasolpropionate-treated areas. Moreover, a significant reduction in nontreated skin sites was observed.

Fig. 5

(a) Exemplary horizontal images of upper stratum granulosum layers by MPT at baseline and (b) after 28 days of treatment at day 29 with characteristic keratohyalingranular morphology.

Fig. 6

Nucleus-to-cytoplasm ratios and cell density counts in the stratum granulosum and stratum spinosum. The nucleus-to-cytoplasm ratio shows an overall increase at day 29 with the highest increase in CP-treated skin sites going along with an increased cell density in all skin sites at day 29. Both nucleus-to-cytoplasm ratio and cell density show partial recovery 7 days after eoT.

3.3.

Cell Nucleus Visibility and Skin Surface Classification

The evaluation of CLSM images showed skin surface properties of abraded skin at baseline measurements in some individuals (Table 5 and Fig. 7). At day 29, an increased number of subjects with properties of abraded skin were observed regarding all four treatment sites. Damaged skin was more frequently observed in steroid-treated skin sites at both day 29 and day 36 with increased frequencies of damaged skin 7 days after eoT. Petrolatum-treated areas also showed an increase in abraded skin and even one case of damaged skin. Cell nuclei were, to a small extent, less than $5/500\mu {\mathrm{m}}^2$ visible on the skin surface at baseline but were clearly distinguishable to more than $5/500\mu {\mathrm{m}}^2$ at day 29 and day 36 in all four skin sites in all subjects indicating parakeratosis or thinning of the stratum corneum layer and enhanced transparency (Fig. 8).

Table 5

Frequencies of the evaluation of the skin surface classification observed at baseline, day 29, and day 36 determined by LSM.

Skin surface classification	CP 0.05%	BDP 0.064%	Petrolatum ointment	Nontreated
	N=8	N=8	N=8	N=8
Day 1/baseline
1 (=healthy skin surface)	7 (88%)	4 (50%)	4 (50%)	7 (88%)
2 (= abraded skin surface)	1 (13%)	4 (50%)	4 (50%)	1 (13%)
3 (= damaged skin surface)	0 (0%)	0 (0%)	0 (0%)	0 (0%)
Day 29
1 (= healthy skin surface)	0 (0%)	1 (13%)	1 (13%)	3 (38%)
2 (= abraded skin surface)	5 (63%)	4 (50%)	7 (88%)	5 (63%)
3 (=damaged skin surface)	3 (38%)	3 (38%)	0 (0%)	0 (0%)
Day 36
1 (= healthy skin surface)	0 (0%)	0 (0%)	1 (13%)	4 (50%)
2 (=abraded skin surface)	3 (38%)	3 (38%)	6 (75%)	4 (50%)
3 (=damaged skin surface)	5 (63%)	5 (63%)	1 (13%)	0 (0%)

Fig. 7

Exemplary images of the skin surface classification observed. In intact skin, the honeycomb pattern with evenly distributed fluorescent signals along the cell borders. Abraded skin exhibits several areas of fluorescent dye accumulation and disturbed honeycomb pattern, while in damaged skin large areas of disrupted skin layers are visible.

Fig. 8

Examples of cell nucleus visibility classification score images by LSM with a visibility of $<5$ cell nuclei, marked by red arrows, within an area of $500\mu \mathrm{m}\times 500\mu \mathrm{m}$ corresponding to class 1 (a) and visibility of $>5$ cell nuclei corresponding to class 2.

3.4.

Effect of GC Treatment on the Collagen and Elastin Signal in the Upper Dermis

Epidermis and dermis differ notably regarding collagen and elastin fiber concentrations and, therefore, in SAAID. For analysis of numerical SAAID values, horizontal multiphoton images were obtained from the upper dermis. Due to the strong reduction in skin thickness, horizontal measurement depths for the SAAID had to be individually adapted to the upper dermis. Therefore, the SAAID were determined at individually determined depths, which was the first measured layer of the upper dermis as the consecutive deeper layer beneath the dermoepidermal junction zone in the horizontal $10\mu \mathrm{m}$ MPT-stacks of two different areas. Despite this analytic procedure, a significant increase of SAAID values was observed in all treatment sites at day 29 compared to baseline with a partial but not full recovery at day 36. Results of SHG measurements showed a corresponding increase of MPT-SHG signal intensities at day 29 in all areas with a higher increase in steroid-treated areas, indicating a higher density of collagen fibers in CP- and BDP-treated skin sites (Fig. 9) with a sixfold increase of MPT-SHG signal intensities found in CP-treated areas, a sevenfold increase in BDP-treated areas, and a twofold increase in both petrolatum and untreated skin areas.

Fig. 9

Mean change in SAAID of the upper dermis at individually determined depths in absolute numbers and percentage from baseline to day 29 assessed by MPT.

4.

Discussion

The reduction in skin thickness observed in this study corresponds to the known atrophic effect of corticosteroids demonstrated in previous studies.^36–38 Moreover, the comparative analysis of skin atrophy using different high-resolution techniques provided more precise in vivo data regarding the different skin layers compared to histological investigations. Here, atrophic structural changes were shown in different skin layers: stratum corneum, epidermis, and dermis. In this study, reduction of skin thickness was also observed in vehicle-treated and nontreated skin areas. The reduction in epidermal and full skin thickness correlates to previous known data about systemic effects of topical applied corticosteroids measured by OCT.¹³ The obtained reduction of stratum corneum thickness after the treatment with petrolatum is on one hand contrary to the known data showing that petrolatum induces a significant swelling of the stratum corneum³⁹ due to its strong occlusional effect.⁴⁰ It is on the other hand explicable by systemic atrophic effects of topical corticosteroids. Petrolatum served as a vehicle basic formulation for both applied corticosteroids in this study. Since the decrease in stratum corneum thickness was observed in petrolatum treated skin, as well as untreated skin, the swelling effect of petrolatum seems to be neglectable. Furthermore, a lateral spreading effect of topically applied steroids is another potential influence factor, which was tried to be avoided by daily application of all ointments by trained medical staff. An abrasive effect by the daily rubbing application of the ointments might have contributed to the increasing number of subjects with signs of abraded skin, since morphological changes of the skin surface by friction can be sensitively detected by LSM without clinical changes of the skin. The high proportion of subjects with damaged skin in GC-treated sites at day 36, 7 days after eoT, although, was unexpected and might indicate a possible rebound-effect, which is well known due to the abrupt cessation of GC treatment.⁴¹

Corresponding to previous in vitro and in vivo studies, keratinocyte size was reduced after steroid treatment in vivo, and this effect was also observed in vehicle-treated and nontreated skin sites. The increased cell density in epidermal layers after steroid treatment seems to contrast with earlier reports on antiproliferative effects of steroids and an overall loss of cells in vitro and in vivo.^36,41–43 It can be assumed that a reduction of the total number of cells occurs with a simultaneously increasing cell density due to the loss of intercellular matrix and ground substance, resulting in an overall reduction of epidermal and dermal tissue volume.

The nucleus-to-cytoplasm ratio has often been used to describe differences in cell morphology. Here, the increase in nucleus-to-cytoplasm-ratio after GC, especially CP treatment, might be an effect of GC-induced atrophic changes.⁴⁴ Corresponding MPT data on an increase in nucleus-to-cytoplasm ratio has been published previously for inflammatory, hyperplastic, or dysplastic skin lesions in comparison to pathohistological sections.²⁹

The cell nucleus visibility was increased in all skin sites at days 29 and 36. This can be explained by the reduction of stratum corneum thickness, enhanced transparency, or structural changes comparable to parakeratosis.

Both, MPT and CLSM provide a high resolution on a cellular level, and CLSM measurements are very suitable to visualize the corneocytes and skin barrier disruptions using the fluorescence mode in combination with a dye. With increasing depth, keratinocyte structures can hardly be seen due to the loss of resolution. Cell nuclei are difficult to see and even cell borders are not always distinguishable. Therefore, cell nuclei and morphologic changes in the epidermal layers, especially in the viable epidermis, can be evaluated with a notably higher precision using MPT.

Numerous in vitro studies showed the reduction in collagen fibers that correlates to the clinical atrophic appearance after long-term steroid treatment.^8,36 A decrease of the SAAID in age-atrophic skin in vivo was shown in previous investigations.²⁴ Here, an increase in collagen signal intensities and a subsequent increase of SAAID values was observed, which seems to contradict these previous data. The findings of this study can be explained by a reorientation of the fibrous network to a more compressed structure. Using electron microscopy, Lehmann et al.⁸ had shown that 6 weeks of occlusive CP treatment lead to a loss in ground substance and reorientation of collagen fibers. While collagen fibers of healthy skin were randomly oriented bundles with separating voids, 6 weeks of occlusive steroid treatment led to a fibrous network with an orientation rather parallel to the skin surface and lacking interfibrous ground substance and voids. The topical application of CP and BDP seems to induce a reorientation of collagen fibers to a more compressed structure, resulting in the loss in total volume explaining the observed increase in collagen concentration and, as a result, in SAAID in the dermis.

Steroid treatment affects the cell metabolism that can induce the increased generation of free radicals. Since free radicals and antioxidants of the skin are at equilibrium, the development of oxidative stress can be seen by the decline in carotenoid concentrations in the epidermis. Here, the steroid treatment itself or mechanical friction might cause the decrease in carotenoids. Furthermore, inflammatory processes especially due to the rebound effect can lead to the generation of free radicals. Moreover, nontreated skin exhibited even increasing antioxidant levels at day 29 and day 36, while both steroid-treated areas showed a significant decrease at day 29 and only partial recovery at day 36. Inferred from the results in nontreated skin at days 29 and 36, it can be assumed that higher amounts of antioxidants are provided by the organism in order to compensate a high radical production in steroid-treated areas as part of a regenerative mechanism of the skin.

The investigation of atrophic, morphological, and antioxidative changes in human skin after steroid treatment was conducted on the volar forearm to avoid additional external influence factors, such as sun exposure, mechanical friction, or occlusion. These results may be comparable to other body parts; skin thicknesses vary depending on body sites, and steroids are known to have stronger effects in intertriginous areas. A confirmatory study including a higher number of subjects and treatment on different skin areas is needed to further investigate the results found in this study.