A Laboratory Method to Measure Skin Surface Staining by Cigarette Smoke, Tobacco Heating Products and E-Cigarettes

All chemicals and reagents were purchased from Sigma-Aldrich (Gillingham, UK, or St Louis, MO, USA) unless otherwise stated.

Four test products were used in this study: 3R4F Kentucky reference cigarettes, iSwitch Maxx e-cigarettes (British American Tobacco, Southampton, UK) and the glo and glo sens THP (British American Tobacco; Table 1). Prior to use, the 3R4F cigarettes were conditioned for a minimum of 48 h and a maximum of 10 days and the glo THP tobacco rods (Neostiks) were conditioned for 48 h to 5 days by storing at 22 ± 1 °C and 60 ± 3% relative humidity, according to the ISO 3402:1999 standard (19). The EC e-liquid cartridges and glo sens THP tobacco pods were stored at room temperature. All devices were fully charged before use.

Ex vivo pig abdominal skin was retrieved in an abattoir and immediately placed on ice. In the laboratory, the skin was prepared by clipping the surface hair and removing excess subcutaneous fat. Slices of 500–750 μm were cut with a dermatome, from which cylindrical biopsy punches of 5 mm diameter were obtained. Prepared skin punch samples were stored at −20 ± 5 ºC until required. On the day of analysis, skin punches were removed from the freezer, brought to room temperature and examined for obvious defects, such as tears, fissures or scratches. Prior to use, skin punches were incubated for 10 min at standard culture conditions (5 ± 1% CO₂ and 37 ± 1 °C) in phenol-free Hanks’ Balanced Salt Solution, containing calcium and magnesium (Thermo Fisher Scientific, Grand Island, NY, USA, 14025-092), and then washed in calcium- and magnesium-free Dulbecco's Phosphate Buffered Saline (PBS) (Thermo Fisher Scientific, 14190-144).

3R4F CS and aerosols from glo, glo sens or iSwitch Maxx were generated using LM20X or LM20E linear smoking machines (Borgwaldt, Hamburg, Germany). The iSwitch Maxx was tested at the highest power level. Specific puffing regimes were used for each product (Table 2). The particulate fraction of CS or aerosol from each product was collected on 44-mm Cambridge filter pads (CFPs, Whatman, Maidstone, UK) and the particulate matter (PM) was eluted with dimethylsulfoxide (DMSO) as described previously (20, 21). Briefly, CFPs were weighed before and after aerosol collection to determine the weight of the captured aerosol. CFPs were placed in 100-mL glass bottles and DMSO added to achieve a concentration of 24 mg/mL. Bottles were covered with foil and placed, at room temperature, on an orbital shaker set at 150 rpm for 25 min. The PM was then extracted from each CFP under vacuum, 1-mL aliquots were prepared and stored in glass vials at −80 °C until required.

Cylindrical punch biopsy skin samples were placed dermis down into 500 μL of product PM or DMSO and incubated under standard culture conditions (5 ± 1% CO₂ and 37 ± 1 °C). Samples were removed at 0.25, 0.5, 1.0, 3.0 and 6.0 h (three skin samples per timepoint) and colorimetric readings were taken. Each experiment was repeated at least three times.

Cylindrical biopsy skin punches were placed into 12-well hanging inserts (Transwell^®, Corning, Lowell, MA, USA) and then into a VITROCELL^® 12/6 CF module (Waldkirch, Germany) with Hanks’ Buffered Salt Solution at ambient room temperature. Using product-specific puffing regimes (Table 2) and a VITROCELL^® VC1^® engine, samples were exposed to 50, 100, 200 or 400 puffs of 3R4F CS, glo sens, iSwitch Maxx or air control (three skin samples per dose). Three or more independent experiments were performed for each product or control, glo was not assessed as an aerosol.

Prior to exposure, the colour profile of each pig skin punch sample was determined using a CM-700d spectrophotometer (Konica Minolta Business Solutions, Greenville, SC, USA) with 5-mm aperture that was calibrated using the manufacturer-supplied white tile before use. Four measurements per skin samples were taken and the sample was rotated 90° between each measurement. Throughout the colorimetric analysis, the operator maintained a uniform specimen measuring port-to-tissue surface distance and ambient lighting to minimize variability and bias in measurements. Colour readings were captured, stored in the CM-700d using the SpectraMagic NX software (Konica Minolta Business Solutions) and the results were exported to a Microsoft Excel document. Samples exposed to PM or DMSO were rinsed in 500 μL phosphate-buffered saline before analysis. Those exposed to aerosol or air were removed from the exposure chamber and placed directly on to the CM-700d spectrophotometer aperture.

Colour profiles and staining levels were calculated at baseline and at every timepoint or puff number using the “Commission Internationale de L’éclairage L*a*b* method”. L* is a measure of lightness and a* and b* are measures of green-red and blue-yellow colour components, respectively (6, 9, 22). Changes in values from baseline and between treatments were determined in Excel by calculating ΔL*, Δa*, Δb* and ΔE (total difference) with the following equation:

The data analysis for this paper was generated using SAS software, Version 9.4 of the SAS System for Windows (Copyright © 2021 SAS Institute Inc., SAS and all other SAS Institute Inc. product or service names are registered trademarks or trademarks of SAS Institute Inc., Cary, NC, USA.) Generalised linear models were used to assess the differences in ΔL*, Δa*, Δb* and ΔE values between the products and reference cigarettes. The significance threshold for difference (α) was set at p = 0.05. Post-hoc Tukey adjustment for pairwise comparisons was also used.

Exposure to 3R4F PM resulted in darkening and discoloration of the skin samples, with effects increasing over time. After 0.25 h, 3R4F ΔL* values were significantly lower than glo sens and DMSO values (p < 0.05, Table 3) indicating darkening of the skin samples. After 0.5 h, 3R4F ΔL* values were also significantly lower than glo and iSwitch Maxx values (p < 0.05). After 0.25 h exposure, 3R4F Δa* values (green to red), were significantly higher than glo and DMSO (p < 0.0001) demonstrating reddening of the skin following the exposure. At 0.5 h, all products and DMSO Δa* values were significantly lower (p < 0.0001) than 3R4F. The Δb* values (blue to yellow) following 3R4F exposure were significantly higher than glo, glo sens and DMSO control from 0.25 h demonstrating that the skin yellows with exposure. From 0.5 h, iSwitch Maxx and all other products Δb* values were significantly lower than 3R4F (p < 0.0001). From 0.25 h, total colour changes, shown by the ΔE value (Figure 1, a), were significantly higher for 3R4F than DMSO and all products except iSwitch. At 0.5 h, glo, glo sens, iSwitch Maxx and DMSO control ΔE values were significantly lower than the 3R4F value (p < 0.0001). All THP and EC values were similar to those for the DMSO control throughout the timepoints assessed (Figure 1, a and b).

Figure 1

Exposure to 3R4F CS aerosol resulted in darkening and discoloration of punch skin samples, with dose-dependent changes observed for ΔL*, Δa* and ΔE* values (Table 4). After 50 puffs, 3R4F CS ΔL* values were significantly lower than those for glo sens, iSwitch Maxx and air control (all p < 0.0001). Skin reddening was seen with 3R4F CS exposure compared with the other products and control, with the difference in Δa* values becoming significant from 50 puffs (all doses p < 0.0001). Skin yellowing, represented by Δb* values, increased following 3R4F exposure, differing significantly from glo sens, iSwitch Maxx and air control values at all puff numbers (all p < 0.0001). However, 3R4F Δb* values reached a plateau at 200 puffs and decreased at 400 puffs. The ΔE values (Figure 1 c), indicating overall colour change, were significantly higher for 3R4F CS than for glo sens, iSwitch Maxx and air control at 50–400 puffs (all p < 0.0001). As for exposure to PM, all THP and EC values were comparable to air control at all doses (Figure 1, c and d).