In vivo transepidermal water loss: Validation of a new multi-sensor open chamber water evaporation system Tewameter TM Hex
INTRODUCTION
Instrumentation technology for transepidermal water loss (TEWL) assessment has not been substantially modified since its introduction
by Nilsson in 1977 when a method for direct measurement of the rate of evaporation from the skin surface has been described.1-4 Classically two sensor pairs for relative humidity (rH) and temperature are placed in an open chamber to detect the concentration difference along increasing distance from the skin. This has been the standard setup for over 40 years now. Variations of this technology are using one sensor pair in a closed chamber.5
Since its introduction, the TEWL has been applied in different fields of dermatological practice and research, as skin barrier phys- iology parameter in healthy and diseased skin, proof of concept for the cosmetic industry, monitoring therapeutic efficacy, occupational aspects of skin disease, and finally as a predictor for asthma and atopic dermatitis development.6-10
The measurement uncertainty of classic probes has physical limits because diffusion is a dynamic process. 11 It is based on the random Brownian motion of air and water molecules. This causes smaller or larger fluctuations already under ideal measurement conditions and even more in challenging situations like outdoor measurements or measurements in a very hot environment. That means there is a phys- ical limit to the precision we can reach when measuring temperature and rH with one sensor.
Modern sensors are designed to address these physical limits. They allow radically new probe designs with advantageous sensor arrange- ments enabling innovative data evaluation techniques and making measurements more precise than with traditional techniques.
Currently, to determine the diffusion rate of water vapor on the skin using the open chamber method, at least two spatially separated pairs of sensors-one thermometer and one rH sensor each-are required to determine the absolute water vapor concentration c in g/m3. The con- centration difference between the two sensors is proportional to the diffusion rate J of the water vapor, which is calculated using Fick's law in the unit g/m2/h.12 The water vapor concentration c is determined in g/m3:
There are different methods for TEWL measurement; the unventi- lated chamber (closed) method like Aquaflux, H4300, and Vapormeter, the ventilated chamber method like MEECO, and the method using an open chamber Tewameter and DermLab.15 The unventilated chamber potentially occludes the skin and thus is incapable of performing con- tinuous measurement over a longer time period. A certain drawback of the ventilated chamber method using dry or moistened carrier gas is the interference with the microenvironment over the skin surface. The open chamber method does not interfere with the microclimate and does not occlude the skin; therefore, it is a reliable and useful tool for both single and continuous measurements of the evaporative loss from the skin surface. 15 A comparison between 3 closed chamber instruments and four open chamber devices was performed in differ- ent models (human skin in vivo, hairless mice in vivo, ex vivo skin model using a gravimetric assay) and across a wide range of perturbations (mild, moderate, and severe barrier damage).11 A high Pearson correlation coefficient was detectable for data from all instruments versus gravimetrically assessed TEWL. Nonetheless, all methods could be influenced by the microclimatic changes near the skin surface. Hence, the measurements must be performed in acclimatized rooms with controlled air temperature and rH without direct airflow into the test field.
The study objective was to compare the new, multi-sensor open chamber water evaporation instrument (TM Hex) for epidermal barrier assessment with the established TM 300 probe. The primary hypoth- esis was that the values assessed with TM Hex and TM 300 at eight different anatomical locations had a significant and relevant correla- tion. The primary endpoint was to calculate the correlation coefficient and whether the possible correlation reached the significance level. Our secondary hypothesis was that the measurement quality (robust- ness) of TM Hex is not inferior to TM 300 in terms of repeatability on the volar forearm. The secondary endpoint was to compare the coefficient of variance proof equivalence or comparability of the established probe TM 300 with the new TM Hex probe. In addition, we introduced new barrier-related parameters assessed with the TM Hex as transepidermal heat loss. Reference values for transepidermal heat losses caused by heat diffusion and by evaporation cooling are reported in a descriptive way without a formal hypothesis. The local water vapor concentration is tested as an indicator of local ambient conditions at the measurement location and their influence on the TEWL measurements.
MATERIAL AND METHODS
Study design
This prospective, comparative, observational study was conducted on 24 healthy Caucasian volunteers (nine male and 15 female) with a mean age of 29.3 (20-55) years. Ten repeated baseline measurements were performed on the volar forearm with the new Tewameter TM Hex and the established TM 300. An additional seven anatomical location measurements were performed with TM Hex. On the volar forearm, additional values of stratum corneum hydration were assessed. Topical treatment (medication) for the last 4 weeks, cosmetic product applica- tion for the last 24 h, and systemic medication (immune modulators, antihistamines, nasal decongestants, and hormones) potentially inter- fering with epidermal barrier function were exclusion criteria. In order to further show that the analyzed skin areas had no further alter- ations SC hydration was assessed with the Corneometer CM 825 and skin color for pigmentation (Melanin index) and inflammation (Ery- thema index) with the Mexamenter MX16. The study was carried out respecting the published guidelines for TEWL, skin hydration, and skin color.16-18 We did not perform a formal power calculation prior to the start of the study. This study was intended to show a non-inferiority of the Tewameter TM Hex we did an estimation of the needed sample size based on published comparative studies.11,19,20
The study was submitted and approved by the Ethical Commit- tee of the Euroderma Clinic in Sofia under the number 01-2020.
Schematic representation of the TM Hex. Panel A: Tewameter Hex cut open with sensors at different distances to the skin in contact with the diffusing water molecules. Panel B: Entire measuring probe with handle.
Measured variables
The sensor matrix provides spatially resolved measurements of rH and temperature. Robust statistical methods are used to model the spatial distribution of temperature T and rH. From these mea- surements, the concentration of CH20 in water vapor can be calcu- lated robustly.21 A large number of sensors enables new ways of evaluating the data, which in turn makes new measurement vari- ables possible. The use of statistical methods significantly reduces the measurement uncertainty and improves the reproducibility of experiments.
Transepidermal water loss
As shown in Figure 2, the concentration gradient in g/m3/h is calcu- lated from the 30 concentration measurements and the distances of the sensors to the skin, considering the current air pressure. The used algorithm e used algorithm allows high robustness against local devi- ations of measured values due to air turbulences as well as in case of possible failure of individual sensors.
FIGURE 2 Water concentration and extrapolation. Water concentration calculated from each of 30 sensor pairs in TM Hex plotted over skin distance (vertical axis). Measured water vapor concentration in the diffusion tube plotted over distance to the skin (black), fitted polynomial 2nd degree (blue line), extrapolated values on the skin (z = 0 mm) and in ambient air (z = 20 mm) (red circles).
Extrapolated measurement values
Since the spatial curves of the measured temperature and rH along the axis of the diffusion cylinder are known, they can be used to extrap- olate measured values derived from these curves to the endpoints of the diffusion cylinder. Thus, a measured value directly on the skin surface and a measured value in the ambient atmosphere can be deter- mined for each variable. In Figure 2, this is shown for the water vapor concentration by the red circled end points of the best-fit line.
Heat loss
The temperature gradient aT/az is calculated from the temperature values of the 30 sensors. Using Fick's law and the appropriate diffusion constant DT, the heat loss due to diffusion-based heat transport can be determined and expressed in W/m2. Multiplying the evaporation enthalpy of water (about 1.3 kJ/g at 25°C) with the TEWL additionally yields the heat dissipated by evaporation cooling of the body at the measuring point. These two measured quantities describe 50% of the heat balance. The heat dissipation by infrared radiation and heat loss by breathing is responsible for the rest.
Skin surface temperature
By extrapolating the temperature signals to the skin surface and the environment, the skin temperature and the ambient temperature can be determined. In practice, however, very long measurement times of over 100 s would be necessary to obtain stable measured values, since the temperature measurement process converges exponentially. To calculate a measured value during the typical measurement period of 20 s, the convergence temperature can be predicted from the temperature curve using a temporal extrapolation model.
Water vapor concentration
The extrapolation of the water vapor concentration to the skin surface and to the environment allows the analysis of the condition of the skin as well as the local environment during the measurement. In contrast to a remotely mounted temperature and humidity sensor that measures room conditions, it is possible to document what conditions prevailed at the measurement site. The difference between skin and ambient conditions is the driving force of the diffusion and thus is proportional to TEWL. A high water concentration in ambiance can, for example, show up when measuring skin areas that have been occluded before starting the measurement.
Measurement uncertainty estimation
Statistical modeling of rH and temperature (T) inside the diffusion tube of TM hex has the benefit, of providing information about the instrument's measurement uncertainty. The measurement uncertainty is elementary information on the quality and validity of measurements and should accompany all measurement values. While this is a well- established standard in industrial metrology22 it is new in the field of skin physiology. The TM hex is the first evaporation probe to provide information on its measurement uncertainty for TEWL.
RESULTS
No adverse events occurred during the study and 24 volunteers com- pleted the study as per protocol. Two volunteers were recruited for training and protocol adjustment purposes. Their data, as planned, did not enter into the statistical analysis. The SC hydration was 42.0 AU (min 29.4-max 66.0; standard deviation [SD] 8.16). The Melanin index showed values of 157.9 AU (90.0-271.0; SD 43.0), and the Erythema index 246.3 AU (145-333.0; SD 52.2). The mean age was 29.9 years (20.0-55.0; SD 8.3). The acclimatized room had a mean temperature of 19.8°C (18.3-21.8; SD 0.75) and a mean rH of 43.7% (39.1-50.9; SD 2.69).
We could demonstrate a significant and relevant correlation between Tewameter TM Hex and the TM 300 (p < 0.0001) mea- sured on the seven anatomical regions with an excellent Spearman R-coefficient = 0.8248. (Figure 3A). In clinical studies, the palms are usually not analyzed due to their high values (based on the pre- dominant sweat gland activity). We performed a second calculation including the high values of the palms, which confirmed the data (Figure 3B); (p < 0.0001; R-coefficient = 0.8778). The repeatability was compared on the volar forearm calculating the SD in g/m2 h and the coefficient of variance (CoV) in % (Figure 4). The SD was signifi- cantly (p < 0.0001) lower for TM Hex (0.8667 g/m2 h) than for TM 300 (1.2380 g/m2 h). Accordingly, the CoV was significantly (p < 0.0001) lower for TM Hex (11.35%) than for TM 300 (21.21%). Furthermore, the CoV varied only slightly between the different anatomical sites between 7% (right cheekbone and right inner upper arm and 14% [right palm]) (Figure 5). In addition, we introduced new barrier-related parameters assessed with the TM Hex. Here we did not work with a formal hypothesis and descriptively report the data. Figure 6 repre- sents the water vapor concentration in air cH2O (calculated at the skin surface) which did not show a relevant correlation with SC hydration (measured in the depth of SC). Figure 7 depicts the transepidermal heat loss on the 8 anatomical locations. The blue bars represent the transepidermal heat loss by heat diffusion (W/m2). The values vary around 7-11 W/m2 for all areas (mean 8.5; SD 1.5). The orange bars depict the transepidermal heat loss by evaporation (W/m2) with com- parable values between 4.9 and 31.2 (palm) with an SD between 1.3 and 17.2 (palms). A linear correlation between heat loss and age could not be found but there is a trend of heat loss rising with age, especially in palms. (See Figure S1) Figure 8 shows the water concentration in the air on the skin and ambiance (conditions that prevailed at the measure- ment site) (Figure 8A) and the rH in the air on the skin (Figure 8B). They showed comparable results for both parameters except for the palm as a measuring point. The high SD of CH2O at the palm confirms the turbulent atmosphere at this location.
We compared the time to reach steady-state conditions for TM Hex and TM 300. The time was approximately 20 s for both probes without a significant difference (Figure S2). This value correlates with the phys- ically determined expected convergence time limited by the speed of water vapor diffusion in the air. The speed is approximately 900 s/m at typical lab conditions, resulting in 18 s needed to cover the 20 mm long diffusion tube of both Tewameter types.
DISCUSSION
TEWL is the gold standard in vivo parameter for the evaluation of epidermal permeability barrier function. TEWL represents insensible
FIGURE 3 Correlation of TM Hex and TM 300. Panel A: Measurements from seven locations excluding palms with a significant correlation of p <0.0001 and a relevant Spearman correlation coefficient of
r = 0.8248. Panel B: Measurements from all
eight locations with a significant correlation of p<0.0001 and a relevant Spearman correlation coefficient of r = 0.8778.
perspiration which is based on the diffusion of body water through the SC. Minimizing thermal sweating, hence, is crucial for quantifying TEWL under basal conditions. A low TEWL is a characteristic feature of a balanced, healthy skin protective state. TEWL measurements can be used to assess the inside-out barrier. TEWL values are indirectly rep- resentative of the outside-in barrier predicting the penetration rate of topically applied substances and pharmaceutical compounds thru the epidermis.
The limitations of classical probes assessing TEWL are the use of only one or two sensors. The development of new sensors allowed the development of a new probe with higher numbers of sensors without compromising the readiness to measure small areas on all anatomical areas of the human body. A new instrument needs to show comparable precision and robustness to established systems.
Our primary hypothesis that TM Hex and TM 300 values cor- relate significantly (p < 0.0001; Spearman correlation coefficient of r = 0.8778) at 8 different anatomical locations was confirmed. The secondary hypothesis that the measurement quality (robustness) of TM Hex is not inferior to TM 300 in terms of repeatability on the volar forearm could be confirmed. We were even able to show that the CV values for the TM Hex are significantly lower than for TM 300 (p< 0.0001). In addition, we could introduce transepidermal heat loss caused by heat diffusion and by evaporation cooling as new barrier-related parameters assessed with the TM Hex. The local water vapor concentration is indicative of local ambient conditions at the measurement location.
The correlation between TM Hex and TM 300 along with the robust- ness of the measurements with TM Hex provide evidence that the new probe for assessment of epidermal barrier function is compara- ble to the TM 300. Most commonly, TM Hex shows even more accurate measurements than the TM 300 probe. The new parameter, transepi- dermal heat loss, allows us to study the local energy balance of skin compared to other heat dissipation mechanisms of the human body, even under extreme conditions. Transepidermal heat loss by heat diffusion (W/m2) and transepidermal heat loss by evaporation have the same order of magnitude as the typical basal metabolic rate of
FIGURE 4
Repeatability: Standard deviation (SD) and coefficient of variance (CV). SD and CV are assessed with the ten repeated measurements on the volar forearm. Dotted lines represent the SD in g/m2 hr and solid lines CV. For both parameters, TM Hex shows significantly lower values than TM 300 except for participants 23 and 25.
FIGURE 5
Repeatability: Coefficient of variance (CV) at different anatomical areas. Repeatability of the measurement is depicted with box plots of coefficient of variance: On the volar forearm direct comparison showed significantly higher CV-values for TM 300 (dark grey box and blue dots) compared to TM Hex (light grey boxes and red dots) (p < 0.0001). The quality of repeated measurements in terms of coefficient of variance ranged between 7% (cheekbone, upper arm) and 14% (palm) at the different anatomical locations for TM Hex. Boxes show the two middle quartiles of each data set.
~1 W/kg body mass. Differences can be explained by other heat dis- sipation mechanisms like breathing and infrared radiation and the inhomogeneous heat dissipation of different skin locations.
Parameters like local rH or water vapor concentration can pro- vide additional information about the experimental situation in non- standard measurement conditions.
The limitation of our study is the medium-sized cohort (n = 24) with only Caucasian volunteers. Furthermore, the study did not address barrier alterations and repair dynamics. Repair mechanism and their modulation are known not to be instrument dependent since each volunteer/patient serves as its control.4.11 Future clinical studies will include the assessment of TEWL values of patients suffering from inflammatory dermatosis with epidermal involvement and the mod- ulation of TEWL values by topical treatments. We expect that TM Hex will be used in studies assessing barrier alterations and their treatments.
In conclusion, we could prove a significant and relevant correlation between TM Hex and TM 300 along with the robustness of the mea- surements with the TM Hex. Our study shows that the new probe for the assessment of epidermal barrier function is comparable to the TM 300. In most conditions, TM Hex provides more accurate measure- ments than TM 300. New parameters open the field to studying skin's water and energy balance.
ACKNOWLEDGMENT
JWF and RD have previously received consulting fees from Courage + Khazaka. GW is employed by Courage + Khazaka.