Evaluating the Irritant Factors of Silicone and Hydrocolloid Skin Contact Adhesives Using Trans-Epidermal Water Loss, Protein Stripping, Erythema, and Ease of Removal

A composite silicone skin adhesive material was designed to improve its water vapor permeability to offer advantages to wearer comfort compared to existing skin adhesive dressings available (including perforated silicone and hydrocolloid products). The chemical and mechanical properties of this novel dressing were analyzed to show that it has a high creep compliance, offering anisotropic elasticity that is likely to place less stress on the skin. A participant study was carried out in which 31 participants wore a novel silicone skin adhesive (Sil2) and a hydrocolloid competitor and were monitored for physiological response to the dressings. Trans-epidermal water loss (TEWL) was measured pre- and postwear to determine impairment of skin barrier function. Sil2 exhibited a higher vapor permeability than the hydrocolloid dressings during wear. Peel strength measurements and dye counter staining of the removed dressings showed that the hydrocolloid had a higher adhesion to the participants’ skin, resulting in a greater removal of proteins from the stratum corneum and a higher pain rating from participants on removal. Once the dressings were removed, TEWL of the participants skin beneath the Sil2 was close to normal in comparison to the hydrocolloid dressings that showed an increase in skin TEWL, indicating that the skin had been highly occluded. Analysis of the skin immediately after removal showed a higher incidence of erythema following application of hydrocolloid dressings (>60%) compared to Sil2, (<30%). In summary, this modified silicone formulation demonstrates superior skin protection properties compared to hydrocolloid dressings and is more suitable for use as a skin adhesive.


Adhesive Dressing Microscopy Analysis
FTIR FTIR analysis of the materials showed the different functional chemistry of the adhesive surface as shown in Figure S1.Assignment of primary peaks in the silicone FTIR spectra was carried out following the work of Salih et al. 1   Silicone: 1080 (stretching vibration Si-O-Si), 1250 (bending vibration Si-CH 3 ), 1450 (rocking vibration -CH 2 -), 1530, 1630, 1710, 2860, 2920, 3330 cm -1 .

SEM
Of all the samples analysed the Sil2 was the most difficult to analyse via SEM as the material underwent a slow expansion when exposed to the vacuum required for imaging (Figure S2).
For this reason sample analysis had to be carried out rapidly before the material expanded into an open foam structure (the surface visibly became rougher and stressed once removed from the chamber) and sections of the material appeared visibly damaged by the electron microscope at high levels of magnification (see Figure S3).Overall the humidity of the Sil2 and the Mepilex were higher than the hydrocolloid equivalents.Increasing the temperature had no significant impact on the vapour permeability measured, although it did increase as the sample ambient humidity was increased.This contradicts the silicone adhesives which showed decreasing vapour permeability with both temperature and relative humidity.

Swelling %
Swelling measurements were carried out on slices of each adhesive dressing by comparing specified temperatures and humidities as shown below in Table S1 and Figure S5 S5

Adhesive Dressing Moisture Management Comparisons
The relative hydrophobicity of the dressings were determined by comparing water contact angle measurements using a VCA Optima (AST Products, USA) camera.A liquid of water was deposited on the dressing and illuminated to determine the contact angle.This was repeated three times and an example image is shown in Figure S6.The contact angle was determined to be 98 ± 5.0 for the hydrocolloid and 112 ± 1.3 for Sil2, showing a clear statistically significant difference between the two materials (p < 0.0001).However both dressing adhesive surfaces demonstrate overall hydrophobic properties -the silicone more than the hydrocolloid but it is not correct to claim one is hydrophilic and the other is hydrophobic.At best the hydrocolloid can be considered closer to neutral.
The relative vapour permeability of the dressings were measured after acclimatisation in a primitive skin model system -with the adhesive flange sandwiched between a PU sheet and Whatman filter paper as shown in Scheme S1.This was done to mimic the use of these adhesives on human skin where the upper PU film mimics the protective layer on top of a medical dressing, and the filter paper separates the adhesive from moisture sources.Below the lower barrier layer a standard piece of tissue paper was used, either wet or dry, in an experimental design created to mimic the skin surface where a high moisture content is separated from the adhesive by a thin dermal barrier.In this way it was intended we could observe changes in the material vapour permeability in the absence / presence of moisture vapour permeating through the system.
Scheme S1 -Adhesive sandwich test where the material is acclimatised at relative temperature and humidity It is interesting to note that before we began this experiment we tested the relative occlusion of the PU Film used in the absence of an adhesive, interrogating how additional layers restricted the TEWL.Here we found that one layer of PU film is relatively permeable, further additions of the film, and that when the system was exposed to humidity a higher vapour loss was observed providing a statistically significant variation (P = 0.0001).
Increasing layers of PU however reduced the effect as the upper layer (through which vapour measurements were carried out) were increasingly separated from the moisture source (average P = 0.0896).The data is shown in Figure S7.
Figure S7 -Vapour Loss through 1 -5 layers of PU when system was dry (orange) and in presence of moisture (blue).
Additionally there was no significant difference between the data measured after 1 or 4 hours acclimatisation (data not presented) and so future measurements were carried out after 1 hour acclimatisation.
The analysis of hydrocolloid and silicone materials was then carried out and produced the data shown in Figure S8.

Staining of Human Fibroblast Cells
Studies were undertaken of both the biocompatibility of the dressings and CC stain against human fibroblast cells.Initial compatibility measurements were carried out using human fibroblast cells donated by F45 (female aged 45 years old).
A visual cell count was performed both before, 1 hour following and 48 hours following 2 ml CC addition to an incubating cell suspension.The cells took up the majority of the dye within 1 hour of staining (Figure S9 and Table S2) in a uniform manner revealing on subcellular structures.Analysis of counted cells shows no statistical difference in cell numbers following staining (p = 0.44).Cells were also cultured on top of both hydrocolloid and silicone medical dressings.
Cell proliferation was measured using a WST-1 assay kit using a transwell plate.500μl of WST 1 was added to each well, incubated at 37 °C with 5 % CO 2 for 4 hours.Absorbance was measured at 450 nm using a plate reader, determining the average absorbance from three repeats.A background reading at 690 nm was taken to subtract noise from the experiment.This data is shown in Figure S10.The final data, as shown in Table S3, indicates that cell proliferation was uninterrupted by the presence of either the silicone or hydrocolloid dressing, although a one-way ANOVA statistical test showed that the statistical difference between the cells alone, with silicone or

S12
with hydrocolloid are not significantly different.This is as expected for a biocompatible product.
Table S3 -Absorption of WST-1 transwell plates following WST-1 Assay.Finally cells cultured on the hydrocolloid and silicone were stained using CC dye and the proliferation of protein on the dressing surface was compared after destaining, as shown in Figure S11.Unlike protein stains (where solutions of bovine serum albumin was added to the dressing) there was greater background adhesion of dye on the material surface, particularly for the hydrocolloid dressing.

TEWL ANOVA Data
The TEWL data from all participants was used in a 2-way ANOVA comparison between 6 and 24 hours are shown in Table S4.

Patch Loss Rate
The study was designed so that across 30 participants there would be 124 patches of Sil2 and hydrocolloid applied respectively, with 62 on forearms and 62 on torso.After the study was complete 101 of 124 silicone patches had remained in place whilst 122 of the hydrocolloids had remained.The loss of Sil2 was found to be 89% retention on the forearms and 75% retention on the torso.
In preparation of this study Sil2 was cut into smaller strips (2 -3 cm square) -and it was observed by participants who lost patches that the slow peeling from the edge led to it detaching entirely with just a few knocks.This implies a correlation between patch size (surface area of adhesion) and sample stability.Unfortunately this was noticed too late for the patch sizes to be increased for the full participant trial to be altered.
Only patches that retained in place to the end of the participant trial are counted in the data presented in the manuscripts, and samples that detached early were noted and then discarded from future analysis.

TEWL Variability
In order to understand the variability of TEWL measurements across an individual sample two volunteers (1M, 1F) recorded their TEWL regularly over a three / two week period, noting the time and their activity levels before each measurement.The data from this was then used to estimate the variability in TEWL depending on environmental conditions.The volunteers both had a clear distinction in the TEWL of their upper arms (M average 6.75 ± 1.33, F average 5.33 ± 1.79) compared to their torso (M: 3.75 ± 0.98, F: 4.73 ± 1.20).This is not a trend unique to this individual and is repeated commonly across participants.Of particular interest to us was the variability / change in TEWL measurements over time.A 2-way ANOVA was carried out on the TEWL data which showed that there was no significant to the date/time that the measurements were taken, or the activity level of the participant beforehand, (P = 0.76) but there was a significant difference between the four body sites (P < 0.0001).T test comparisons of the arms (LA and RA) showed no significant in the datasets (P = 0.3) whilst comparing arms to torso showed significant differences in the male but not the female (M: P<0.001, F: P ≈ 0.04).This indicates that whilst forearm and torso skin can be significantly different both are far more variable within the individual depending on factors unrecorded in the study.

Figure S1 -
Figure S1 -FTIR Spectra of adhesive dressings of A) Silicone Adhesives and B) Hydrocolloid adhesives.

Figure S2 -Figure S3 -Figure S4 -
Figure S2 -Exemplar photographs of SEM.Left: novel silica expansion (bubbling) under vacuum.Right: damage to surface after exposure during electron microscopy analysis

Figure S8 -
Figure S8-Analysis of Silicone and Hydrocolloid adhesive dressings vapour permeability when dry (orange) and in presence of moisture (blue).

Figure S9 -
Figure S9 -images show fibroblast cells before and after being placed in a CC solution and left in an incubator.Incubator was set to 37 °C and 5% CO2.(A) fibroblast cells adhered onto cell culture flask before staining with CC. (B) fibroblast cells left in 2ml of CC for an hour in the incubator.(C) Fibroblast cells left in 2ml of CC solution for 48 hours in the incubator.

Figure S11
Figure S11 Images of human fibroblast cells on top of hydrocolloid and silicon medical dressings and stained with colloidal Coomassie under light microscope (x100).

Figure S14 -
Figure S14-TEWL variability data for A) 1M and B) 1F volunteers across 3 weeks.Top: TEWL measurements across 3 week period, Bottom: Average and std.deviation of TEWL in each body part.

Table S1 -
Measurements of Swelling % measurement of dressing materials

Table S2 -
Fibroblast cell counts from FigureS10

Table S4 .
Summary of TEWL Data Across Entire Participant Study