1 INTRODUCTION

Skin cancer is the most common cancer worldwide and incidences have been increasing steadily over the last decades [1, 2]. Beside seeking shade and wearing protective clothing, applying sunscreen products is an important strategy of photoprotection [3]. The efficacy of a sunscreen product is dependent on the type and amount of UV-filters [4], the vehicle of the product [5, 6], as well as its quality to create a continuous film over the surface of the skin [7]. Commercial sunscreen formulations are constantly being adapted to the state of research and customized to target susceptible groups like children or dermatological patients.

The sun protection factor (SPF) value quantifies the degree of protection of an individual sun protection product. It is a relative measure of how much longer a person can stay in the sun before developing an erythema response to the solar ultraviolet (UV) rays. For each new sunscreen product, the SPF value must be tested. Guidelines, such as those from the International Organization for Standardization (ISO) define an in vivo procedure, in which the ratio between the UV dose leading to a minimal erythemal response (MED) on protected skin and on unprotected skin, is calculated with a minimum of 10 voluntary test subjects [8].

Consequently, the current procedure requires that the tested volunteers are being intentionally exposed to UV irradiation, raising ethical concerns. Sufficient evidence shows that UV radiation is the main environmental cause of non-melanoma skin cancer [9]. It is considered to be the main environmental cause of cutaneous melanoma, too [10]. Therefore, the European Commission for safety of non-food products has advised the development of a non-invasive method for the determination of SPFs [11].

In the past decades, efforts have been made to develop an in vitro procedure: However, no approach measuring the transmission of the sunscreen on roughened polymethylmethacrylate (PMMA) plates has been broadly accepted, yet [12, 13]. While the shape of the absorption curve represents the effectiveness of the sunscreen filters reliably, the magnitude of the absorbance is often either over- or underestimated, because artificial substrates cannot imitate the interaction between sunscreen formulations and skin [5, 14, 15].

Therefore, a method is required, which allows measurements directly on the human skin without inducing damage due to irradiation.

SPF determination via hybrid diffuse reflectance spectroscopy (HDRS) fulfills these criteria.

By comparing the skin's remitted light before and after the application of a sunscreen, the absolute UVA absorption spectrum can be evaluated. Due to high light extinction in the UVB range, the shape of the UVB absorbance spectrum can be measured using an in vitro transmission spectrum and then adjusted to the in vivo diffuse reflectance spectroscopy (DRS) measurements. Different approaches using two monochromators or a polychromatic light source showed a good correlation with conventional SPF testing [16-18].

In the following, a new DRS device for SPF and UVA-PF measurements is presented that uses an LED-based light source, thus providing an adjustable and compact test system, since LEDs can be pulsed, and the UVA and the UVB range can be individually adapted.

2 MATERIALS AND METHOD

2.3 Hybrid diffuse reflectance spectroscopy procedure

With the resulting DRS signals, the transmission can be calculated as according to Equation (1), where R₀ is the diffuse reflectance after product application and R₀ the diffuse reflectance before product application.

()

The in vitro data were provided by the sunscreen manufacturers. The transmission values of the in vivo DRS measurement T_DRS(λ) are divided by the preexposure in vitro value at each wavelength between 320 and 330 nm, to determine the scaling factor C_HDRS'_λ at each wavelength. The scaling value C_HDRS'₀ is the arithmetic mean of these 11 C_HDRS'_λ values. The values of T_PMMA(λ) from 290 to 320 nm are multiplied by the scaling value C_HDRS'₀ and combined with T_DRS(λ) values from 321 to 400 nm to form the transmission spectrum T_HDRS(λ) from 290 to 400 nm (T_HDRS(λ)_290-400nm). Figure 1 shows representative spectra of different stages of the HDRS procedure.

In order to take the photostability of the test product into account (photodegradation), the transmission values of the postexposure in vitro measurements T'_PMMA(λ) are divided by the preexposure in vitro values T_PMMA(λ) at each wavelength between 290 and 400 nm. This results in the wavelength-dependent scaling factor of photodegradation C_PD'(λ).

()

T_HDRS(λ)_{290–400 nm} is multiplied by C_PD'(λ) before calculating SPF_HDRS according to Equation (2), where E(λ) = erythemal action spectrum, I(λ) = spectral irradiance, T'(λ) = postexposure transmission T_HDRS(λ)_{290–400 nm}, and d(λ) = wavelength interval.

We decided to apply the HDRS principle to the in vivo and the preexposure in vitro spectra. Alternatively, one could first correct the in vivo spectra for photodegradation and then use the postexposure in vitro spectra to obtain the HDRS spectra, see Table 2.

3 RESULTS

The SPF and UVA protection factor (UVA-PF) values obtained by this new approach were compared with in vivo SPF results determined by certified test institutes, as illustrated in Figures 2 and 3. A linear regression through zero with a slope of 1.43 and a correlation coefficient R² = 0.86 for SPF and a slope of 1.29 and R² = 0.92 for UVA-PF was obtained. The error bars of the reference values correspond to the 17% maximum standard error, which is allowed according to ISO 24444 [8].

The sunscreens included in the evaluation with their reference, SPF_HDRS-LED and UVA-PF_HDRS-LED values are shown in Table 1.

In addition to the SPF_HDRS values, in vivo DRS data obtained by the multi-lambda device were compared with results determined by test institutes. Taking the photostability into account, an R² = 0.78 was calculated for sunscreen products up to SPF 26 (n = 4), as shown in Figure 4. When considering only test subjects with skin photo type I (n = 2–4), an R² = 0.9 was calculated for sunscreen products with SPF 26 and lower.

In addition to the SPF_HDRS values, in vivo DRS data obtained by the multi-lambda device were compared with results determined by test institutes. Taking the photostability into account, an R² = 0.78 was calculated for sunscreen products up to SPF 26 (n = 4), as shown in Figure 4. When considering only test subjects with skin photo type I (n = 2–4), an R² = 0.9 was calculated for sunscreen products with SPF 26 and lower.

4 DISCUSSION

During the project implementation, various approaches to apply the HDRS principle have been examined. Each of these approaches used a scaling factor C'_λ of each wavelength either between 320 and 330 nm or between 340 and 350 nm to adjust the in vitro spectrum to the in vivo DRS measurements. Different approaches for aging (correction for photodegradation) were investigated. They included no aging at all, calculating the ratio of photodegradation (RPD) based on Rohr et al. [17] and determining a wavelength dependent scaling factor. The correction for photodegradation was done after application of the HDRS principle and alternatively before calculating a HRDS spectrum. In addition, both versions were carried out using either transmission or absorption values for all the calculation steps. The correlation coefficients for all of these options are shown in Table 2.

The method presented above was found to produce the most valid and best reproducible results, which indicates that the multi-lambda device is suitable for SPF testing based on the HDRS principle as well as for UVA-PF measurements. An external device for testing photostability is required, yet.

Compared to waiting 24 hours for erythema formation using conventional SPF determination, this approach is much faster: Measuring one SPF takes about 15 minutes per volunteer plus the time of application and incubation (independent of the formulation).

With an effective dose of less than 20% of the MED of skin type I the technology is non-invasive. No erythemal response is being induced. This makes MED reading, a soft skill often discussed as an inter-grader variable factor of SPF testing, obsolete [22, 23].

The spectroscopic measurements are independent of a biological response and thus not influenced by anti-inflammatory components of certain commercial sunscreen formulations, which can otherwise lead to an overestimation of SPF values [24].

The compact LED-based setup offers a high flexibility in terms of spectrum, light intensity and time. The duration of illumination can be adapted to measurement purposes. Each LED can be individually adjusted in order to tune and to omit specific wavelength ranges. This could be beneficial, for example, to investigate particular wavelength-dependent properties of sunscreen products.

Apart from SPF values based on the HDRS principle, the multi-lambda-LED device produced DRS in vivo values that showed a good accuracy in sunscreen products with SPF 26 and lower, especially when testing subjects with skin type I. This is in accordance with previous experiments using an older Xenon lamp-based spectroscopic DRS setup, where the comparably high reflectance of pig ear skin was mostly in good agreement with SPF values [25]. Pure in vivo measurements were also shown for a setup using one UVB LED and a photodiode for detection [26]. Now the multi-lambda-LED light source with 8 LEDs covering the whole UV range allows more spectrally resolved and thus accurate measurements. These could be beneficial for testing the reproducibility of SPF values, for example, to determine the sweat-resistance or homogeneity of a sunscreen product [7, 27]. Furthermore, the in vivo spectra may be useful for research purposes, since skin filter interactions are measured in the UVA and in the UVB range. For sunscreen products with SPF values higher than 26 the remitted light signal becomes insufficient for the determination of SPFs. In order to pursue a non-invasive, pure in vivo method to determine high SPF values, too, a modified, more powerful multi-lambda-LED device with a cooling system will be developed.

5 CONCLUSION

Overall, we present a compact and adjustable measurement system using a multi-lambda-LED light source that is suitable for non-invasive and fast SPF testing based on the HDRS principle. In addition, DRS in vivo measurements can be performed and used for research purposes.

CONFLICT OF INTEREST

G.W., C.R., and G.K. are employed by Courage + Khazaka electronic GmbH, which is the industrial partner of the joint research project with an interest in commercialization.

AUTHOR CONTRIBUTIONS

Georg Wiora, Carina Reble, Carolin Maria Throm, Martina C. Meinke and Jürgen Lademann were involved in conceptualization. Carina Reble, Georg Wiora and Georg Khazaka, construction and building up the system. Georg Wiora, Carolin Maria Throm and Johannes Schleusener were involved analyzing the data. Carolin Maria Throm, Sabine Schanzer, Georg Wiora, Carina Reble, Johannes Schleusener, Ludger Kolbe and Hans Karrer were involved in investigation. Carolin Maria Throm writing-original draft; Carina Reble, Johannes Schleusener, Martina C. Meinke and Jürgen Lademann writing-review and editing. Jürgen Lademann, Martina C. Meinke, Georg Khazaka and Georg Wiora, project management.