Clinical, biophysical and immunohistochemical analysis of skin reactions to acute skin barrier disruption - a comparative trial between participants with sensitive skin and those with nonsensitive skin

Dear Editor, One out of two women and one out of three men in the Western world reports having sensitive skin (SS).1 2 3 4 5 Individuals classifying themselves as having SS frequently describe stinging, burning or itching sensations of the skin,6 7 8 and erythema and dryness may accompany these sensations.5 7 8 9 10 As a response to this, unravelling the pathophysiology of this condition and developing solutions for individuals with SS has become an important topic in biomedical and cosmetic research fields. Although many clinical evaluations and biophysical techniques have been applied in order to shape understanding of and create consensus on the causal pathways of SS, no diagnostic test has been found to be sufficiently sensitive and reproducible. The prevalence of SS and burden on dermatologists stress the urgent need for recognition of SS, and the necessity of gaining more knowledge to reach a consensus on its pathomechanism. Previous studies proposed potential pathways that may lead to a condition commonly designated as SS: impaired barrier function, sensory hyper‐reactivity, inflammatory or vascular responsiveness and atopic predisposition. Unfortunately, studies show inconsistent outcomes, impeding identification of key pathways in the mechanism of SS. In explorative studies, many provocations used as a diagnostic method for SS have resulted in sensory skin reactions10 11 12 13; however, susceptibility to one provocative agent does not predict susceptibility to another.14 15 16 17 In this study, we aimed to identify key features in the morphology and physiology of skin reactions and skin recovery in participants with SS. We evaluated skin reactions to standardized provocation using an in vivo model for barrier disruption in selected participants. Evaluation was performed dynamically, that is at baseline and at several time points after acute barrier disruption. As tape stripping is widely used as a reproducible in vivo human model of cutaneous injury and regeneration, this allows the study of SS in a dynamic fashion.18 In order to reduce confounding by concomitant skin conditions, we propose an experimental design with exclusion of skin diseases. We are of the opinion that a multifactorial questionnaire encompassing a range of provocations, including chemical, mechanical and thermal origins, and multiple signs and symptoms could be a more adequate selection tool than topical agents only. These methods, combining self‐reported skin sensitivity with biophysical and detailed histological evaluation of skin responses, could enable the exploration of a range of processes and delivery of decisive answers to questions on causal pathways of SS, and explain previously described inconsistencies. Participants were recruited via student websites. Volunteers filled out a questionnaire (Appendix S1; see supporting information) on self‐assessed skin sensitivity and skin perceptions (i.e. discomfort, stinging, redness or dryness) following potential endogenous and exogenous triggers (i.e. toiletries, shaving, emotions, heat, cold and clothes). The severity of reactions was scored on a visual analogue scale (VAS), and duration of reactions on an ordinal scale [i.e. no experience, seconds, minutes, < 1 h, hour(s), day(s)]. The investigator scored the 24 VAS scores and 24 durations (recoded to scores of 0, 1, 2, 4, 7 and 10, respectively) of the questionnaire. The upper and lower quartiles of the sumscore of these 48 items (range 0–480) was determined in a large cohort study (n = 238, manuscript in preparation). A participant in the present study was defined as having SS if the sumscore was above the upper quartile value and the patient reported SS. A participant was defined as having nonsensitive skin (NSS) if the sumscore was below the lower quartile value and the patient reported NSS. Participants were included if the following criteria were met: had SS or NSS according to the procedure described above; had Fitzpatrick skin type II or III; did not have or have had an atopic condition [asthma, allergic rhinoconjunctivitis or atopic dermatitis (AD)]; was able to give written informed consent; had no skin disease, including contact dermatitis. Participants were advised not to apply cosmetics for 24 h before the first visit. Groups with SS and NSS skin were sex matched. The experiments were approved by the local ethics committee (METC, region Arnhem‐Nijmegen) and were conducted to conform with the principles of the Declaration of Helsinki. Sequential tape stripping of four skin sites parallel to the upper side of the intergluteal cleft was performed using a metal oblong plate with an oval aperture (13 × 22 mm, 2·9 cm2) covered by 6890 polyvinylchloride tape (Scotch® Tape; 3M, St Paul, MN, U.S.A.) to standardize extension of the skin and velocity and angle of removal.18 19 Stripping was stopped when the skin became homogeneously refulgent. Macroscopic images were taken as a reference and tapes were counted. Participants were acclimatized to room conditions for 20–25 min (21 ± 1 °C, 50 ± 10% relative humidity). Measurements were performed three times on nonhairy, optically healthy skin, without cleansing the skin, in the following order. (i) Skin redness (a*) was measured using a spectrophotometer (Chroma Meter 2600d; Konica Minolta, Tokyo, Japan). Settings used were as follows: 8 mm changeable aperture; 100% ultraviolet (UV; illumination that contained all UV components of xenon flashlight source); M/I+E [specular component excluded (SCE) was used as the readout setting]; 65 d illuminant (standard illuminant daylight); and a delay time of 0·0 s. The interval between each measurement was 30 s and the device had only brief skin contact, to prevent vasoconstriction and subsequent hyperaemia. (ii) Transepidermal water loss (TEWL) was measured using an AquaFlux AF200 (Biox Systems, London, U.K.) perpendicular to the skin. We used intervals of 2 min, to ensure evaporation of surface water caused by occlusion and condensation. (iii) Stratum corneum (SC) hydration was measured using the Corneometer CM825 (Courage+Khazaka Electronic GmbH, Cologne, Germany). Intervals of 10 s were applied. Prior to the biopsy procedure, skin was cleansed with chlorhexidine 0·5% w/v in alcohol 70% v/v. Punch biopsies (3 mm) were taken at baseline and 0·5, 8, 24 and 72 h after tape stripping under 1% xylocaine–adrenaline.18 19 The tissue samples were fixed in 10% formalin for paraffin embedding, and sectioned at a thickness of 6 μm. Paraffin‐fixed specimens were deparaffinized and rehydrated. Antihuman primary antibodies used are summarized in Table 1. Antigen retrieval was achieved by citrate buffer (pH 6·0, 10 min at 100 °C) for Ki67, CD1a, CD31, K16 and filaggrin staining. To retrieve the epitopes for the tryptase antibodies and CD3 antibodies, TRIS/ethylenediaminetetraacetic acid (EDTA) (50 mmol L−1 TRIS + 2 mmol L−1 EDTA, pH 9·0, 10 min at 100 °C) and EDTA/Tween‐20 (10 mmol L−1 EDTA + 0·05% Tween‐20, pH 8·0, 10 min at 100 °C) were used, respectively. The S100 and elastase primary antibodies did not require antibody retrieval steps. Endogenous peroxidase activity was blocked by 3% H2O2 in phosphate‐buffered saline for Ki67, CD31, S100, CD3 and tryptase antibodies, and preincubated in 1% bovine serum albumin (BSA; Dako, Glostrup, Denmark) for 15 min before application of Ki67, CD31, K16, S100 and tryptase antibodies, and 30 min before CD3 antibodies. Before antifilaggrin application, sections were preincubated with 20% normal horse serum (Vector Laboratories, Burlingame, CA, U.S.A.). Sections were incubated with primary antibodies dissolved in 1% BSA overnight at room temperature; filaggrin and K16 antibodies were incubated for 60 min. Amplification was obtained by horseradish peroxidase‐labelled polymer (EnVision antirabbit; Dako) for S100, and by EnVision antimouse (Dako) for the other antibodies. 3,3′‐Diaminobenzidine tetrahydrochloride solution (Sigma‐Aldrich, St. Louis, MO, U.S.A.) was used to visualize the antibody. As a counterstain, Mayer's haematoxylin (Sigma‐Aldrich) was used. Finally, slides were dehydrated and mounted using Permount (Thermo Fisher Scientific, Waltham, MA, U.S.A.). Filaggrin and K16 antibodies were visualized with Alexa 549 (red) 1 : 200 and Alexa 488 (green) 1 : 200 (Invitrogen, Carlsbad, CA, U.S.A), respectively. 4,6′‐Diamidino‐2‐phenylindole 1 mg μL−1 1 : 3000 (Vector Laboratories) was used as a counterstain. Antibody specifications aDako, Glostrup, Denmark. bAbcam, Cambridge, U.K. cSanbio BV, Uden, the Netherlands. dCovance, Richmond, VA, U.S.A. Antibody specifications aDako, Glostrup, Denmark. bAbcam, Cambridge, U.K. cSanbio BV, Uden, the Netherlands. dCovance, Richmond, VA, U.S.A. Quantification was performed using an Axiokop 2 MOT microscope, an Axiocam MRc5 digital camera and AxioVision software release 4.8 (all Zeiss, Jena, Germany). Epidermal thickness was calculated by dividing the epidermal area by the average of the basal membrane (BM) length and the surface length. Semiquantitative analysis was performed using macrocolour deconvolution and colour threshold in ImageJ version 1.46 (National Institutes of Health, Bethesda, MD, U.S.A.). Ki67‐positive nuclei were expressed as the number per mm BM. Filaggrin was expressed as the number of positive pixels per mm surface length. K16, S100 and CD1a were expressed as the percentage of the viable epidermis. CD31, CD3, elastase and tryptase were expressed as the number or area of cells per mm2 dermis, from the BM down to 300 μm. Spongiosis was assessed by two blinded observers (0, no spongiosis; 1, slight spongiosis in a limited area; 2, slight spongiosis in a large area; 3, strong spongiosis in a large area) and consensus was achieved by discussion. The characteristics are presented as median (range). Student's t‐test was used to analyse differences between skin sensitivity groups (self‐assessed SS, self‐assessed NSS) for continuous variables at baseline. A linear mixed model was used to study differences between the sensitive and nonsensitive groups. The dependent variable was the specific variable: biophysical measurements (TEWL, SC hydration and a* value) and immunohistochemical quantifications (thickness of epidermis and SC, spongiosis, CD1a, CD3, CD31, elastase, filaggrin, K16, Ki67, S100 and tryptase). The independent class variables were skin sensitivity and time (baseline, 0·5, 8, 24 and 72 h after stimulus) and the value of the specific variable at baseline. The estimated mean differences between groups with 95% confidence intervals (CIs) are presented at each point of measurement. Analyses were performed using SAS version 9.2 for Windows (SAS Institute, Cary, NC, U.S.A.) and IBM SPSS 20.0 (IBM, Armonk, NY, U.S.A.). P‐values < 0·05 were considered statistically significant. In both groups (SS group, n = 8; NSS group, n = 8), sex was equally distributed (Table 2). Thirty‐seven participants were excluded owing to the presence of an atopic predisposition, and 20 were excluded as they did not have a score matching their skin sensitivity (either too high or too low). Participants with SS and those with NSS had a median age of 22·5 years (range 20·0–28·0) and 20·0 years (range 19·0–24·0), respectively. The median questionnaire scores of the SS and NSS groups were 150·6 (range 68·0–363·0) and 30·8 (11·0–44·0), respectively. Four participants with SS reported dry skin of the body and two with SS showed a flare and weal reaction following a dermography test, while none of the NSS participants reported these features. The number of tapes required to strip the SC off was significantly lower in the SS group [24·4 (range 16·8–33·0)] compared with the NSS group [36·8 (range 25·3–57·5)] tapes (P = 0·02). Baseline characteristics of participants with nonsensitive skin (NSS) and those with sensitive skin (SS) Data are n (%) unless otherwise specified. Baseline characteristics of participants with nonsensitive skin (NSS) and those with sensitive skin (SS) Data are n (%) unless otherwise specified. At baseline, no significant differences were found between the SS and NSS groups with respect to TEWL (SS: 11·1 gm−2 h−1, NSS: 11·4 gm−2 h−1); SC hydration [SS: 23·0 arbitrary units (AU), NSS: 20·2 AU]; and a* value (SS: 6·0 AU, NSS: 6·4 AU) (Table 3). Observed medians (range) of biophysical measurements and immunohistochemical markers by skin sensitivity and the estimated mean differences at each point of measurement, using a linear mixed model for repeated measurements adjusted for baseline values Arbitrary units (AU) for stratum corneum (SC) hydration from 0 (no water at all) to 120 (water); for a* value colour range from negative (green) to positive (red). CD1a, marker for Langerhans cells; CD3, marker for T cells; CD31, marker for endothelial cells; elastase, marker for polymorphonuclear granulocytes; K16, marker for abnormal differentiation of keratinocytes; Ki67, marker for proliferating keratinocytes; NA, not applicable as these data were not expected to contribute to our research question; NSS, participants with self‐assessed nonsensitive skin; S100, marker for dendritic cells; SS, participants with self‐assessed sensitive skin; TEWL, transepidermal water loss; tryptase, marker for mast cell protease tryptase. aMedian (range). bmean (95% confidence interval). cn = 7. Observed medians (range) of biophysical measurements and immunohistochemical markers by skin sensitivity and the estimated mean differences at each point of measurement, using a linear mixed model for repeated measurements adjusted for baseline values Arbitrary units (AU) for stratum corneum (SC) hydration from 0 (no water at all) to 120 (water); for a* value colour range from negative (green) to positive (red). CD1a, marker for Langerhans cells; CD3, marker for T cells; CD31, marker for endothelial cells; elastase, marker for polymorphonuclear granulocytes; K16, marker for abnormal differentiation of keratinocytes; Ki67, marker for proliferating keratinocytes; NA, not applicable as these data were not expected to contribute to our research question; NSS, participants with self‐assessed nonsensitive skin; S100, marker for dendritic cells; SS, participants with self‐assessed sensitive skin; TEWL, transepidermal water loss; tryptase, marker for mast cell protease tryptase. aMedian (range). bmean (95% confidence interval). cn = 7. At baseline, SC and viable epidermal thickness measured in histology appeared to be comparable between those with SS and those with NSS. The median SC thickness was 15·7 μm (range 10·0–23·0) in participants with SS and 15·6 μm (range 10·3–20·5) in those with NSS, and viable epidermal thickness was 75·5 μm (range 65·6–95·54) and 77·6 μm (70·7–89·8), respectively. Thickness of the granular layer expressing filaggrin in participants with SS and NSS was 5·1 μm (range 2·3–8·5) and 5·6 μm (range 2·5–9·1), respectively. No significant difference between those with SS and those with NSS was found in K16 expression or the number of proliferating cells (Ki67) per mm BM. With respect to mast cell densities, 45·5 (range 39·0–63·5) and 57·4 (range 42·5–69·6) cells were counted per mm² dermis in the SS and NSS groups, respectively (P = 0·10). No polymorphonuclear were in with SS in those with NSS, and (range and (range of the area was for No significant difference was found in the number of S100 + T cells or cells in the epidermis. participants with SS showed a response tape stripping while none of those with NSS a reaction (P < The response did not with skin type (P = A response after one tape strip was in participants with SS and in three with NSS. assessed erythema at or 8 h after participants in each group showed erythema and two in each group showed slight h after participants with SS and two with NSS reported discomfort, burning sensations h after These sensations in two of participants with SS 8 h after after only two participants with SS and only one with SS at 72 h after encompassing and Macroscopic images of the skin at baseline, 1 min, h, 8 h, 24 h and 72 h after With respect to a* TEWL and SC a significant compared with baseline was h after values for a* values and SC hydration were h after and for TEWL or 8 h after biophysical measurements showed a significantly lower a* value in the SS group h after (P = Table 3). the SC resulted in of the viable epidermis 2). after the estimated mean thickness of the viable epidermis was (P < in those with SS than in those with NSS difference 95% the estimated number of is not significantly after 72 h, the viable epidermal thickness and was comparable while the estimated number of was significantly lower in the SS group difference 95% to < In both groups a relative in was 24 h after while 72 h after the number of 2). With respect to spongiosis the estimated was 24 95% and 72 h 95% after = in participants with SS compared with those with NSS. No difference between those with SS and NSS could be with to K16 expression and values of the epidermal in with self‐assessed nonsensitive skin and in with self‐assessed sensitive skin at baseline and at each point of measurement after for thickness of the viable number of proliferating cells per mm membrane and spongiosis score on scale The the < < Spongiosis in haematoxylin and tissue in a patient with self‐assessed sensitive skin at 24 h after × images of for cells (green) and filaggrin (red) at baseline, 24 h and 72 h after was used as a no K16 expression and filaggrin is expressed in the granular 24 h after K16 is expressed and filaggrin is by tape As a positive at 24 h, for filaggrin be after K16 expression be and filaggrin is expressed more in the granular the filaggrin layer The estimated area of S100 cells per mm2 of the viable epidermis was lower in participants with SS (P = and 8 h after (P < difference (95% to and (95% to The number of mast cells time and was lower at all time points in those with SS, and significant 8 h (P = and 24 h after (P = No differences in T endothelial cells or Langerhans cells could be To identify key in the of SS, we used tape stripping as an in vivo model for skin study showed that tapes were required to strip off the SC in those with SS than in those with NSS, while the SC was of comparable a response was only in of participants with SS, while was not in those with NSS. 24 h after the estimated mean thickness of the viable epidermis was significantly in those with SS compared with those with NSS, with significantly in participants with SS. The significantly spongiosis response in those with SS explain the in the viable epidermal thickness in participants with SS. The differences that the questionnaire be an adequate tool with to for research participants with SS and NSS. The lower number of required to strip off the SC in those with SS could be owing to in of In with our previously proposed that the average area of obtained by tape stripping is a more method for self‐assessed SS than TEWL or the acid to baseline TEWL were in an SS In with this, a = < between and skin a skin in to a in the pathomechanism. The strong spongiosis in the SS group explain the relative of cells per mm BM at 24 h, has been in other Spongiosis the lower a* values measured at several time points in participants with 24 One may on to of the SS. The of SS, for skin barrier vascular and atopic predisposition, could not all be in our the of impaired and the spongiosis and flare response that was in those with SS may to vascular other than a mechanical one other pathways in the pathomechanism. in our study did not differences between participants with SS and NSS, and did not parallel the with to the skin lower baseline SC hydration values were in one baseline values have been TEWL at baseline is to be in those with 30 A of the present study is the number of included and the analysis of four in is of skin on the lower to is has been proposed that SS symptoms have with contact dermatitis and these were excluded from our we found strong of spongiosis following tape that this model be of for the in In the differences between these groups of selected that the used questionnaire is an adequate tool with to participants for SS The differences as in those with SS and NSS is in the of the SC and spongiosis to the analysis of the skin barrier as and and vascular could the of SS. the the research the the Netherlands. The to this study were by of is an of and a for this