j. Cosmet. Sci., 52, 91-102 (March/April 2001) Skin optics revisited by in vivo confocal microscopy: Melanin and sun exposure PIERRE CORCUFF, CIdLINE CHAUSSEPIED, GENEVIEVE MADRY, and CHRISTOPHE HADJUR, L'Ordal Recherche, Laboratoires de Recherche Avancde, 93601 Aulnay-sous-bois, France. Accepted for publication February 28, 2001. Presented at the XXI t3 International IFSCC Congress, Berlin, 2000. Synopsis A new confocal prototype dedicated to the exploration of in vivo human skin has been constructed around a laser confocal module (Oz Noran, Inc.) and a skin contact device, assuring perfect stability of skin images. The power of the Argon/Krypton laser source has been limited to 2row to secure safety, and the laser provides three visible wavelengths: 488, 568, and 647 nm. Optical sections were digitized at video rate, providing easy and rapid measurements of the thickness of epidermal layers and time-resolved information. Unexpected details of the epidermis were recorded with the blue laser line. Melanin provided strong reflection of the basal keratinocytes instead of the absorption expected. The 3D reconstruction of the melanin cap in basal keratinocytes confirmed the behavior of melanosomes acting as myriads of nanomirrors that reflected light. Confocal images of the posterior aspect of the forearm were recorded before sun exposure and then for one month after exposure. There was a 25% increase in the thickness of the stratum corneum. Bright inclusions into the dark nucleus of numerous spinous cells were interpreted as local condensation of chromatin. Numerous bright intercellular filaments were attributed to melanosomes filling up dendrites of melanocytes. A striking observation concerned the lack of melanosome caps in basal keratinocytes. In vivo confocal microscopy affords new insight to the role of melanin and its gradual migration after sun exposure. INTRODUCTION The way by which light penetrates through skin tissue remains a complex phenomenon due to the multiple stratification layers as well as the biochemical composition of the epidermal and derreal layers. It has been established that skin optics respects the rules of absorption, reflection, and scattering (1-3). Absorption coefficients have been experi- mentally determined and mainly depend upon the content of melanin in the epidermis and the blood perfusion (hemoglobin) in the dermis. They rapidly decrease at longer wavelengths. Reflection corresponds to the index mismatch at various interfaces, i.e., air/skin surface, water/protein, etc. (4). Reduced scattering involves the heterogeneous organization of the dermis and the anisotropy of the collagen network. It follows the Raleigh limit, exhibiting the well-known K-4 behavior (5). In such a context, in vivo 91

92 JOURNAL OF COSMETIC SCIENCE confocal microscopy uses a reflected signal to form an image whose contrast is improved by strong reflection and by a limited lack of photons due to absorption and scattering. Moreover, confocal microscopy of the skin requires real-time imaging to limit blurring caused by motion, i.e., blood flow pulses and involuntary movement. Until recently, the tandem scanning microscope (TSM) was the only technique having this prerequisite of real-time imaging, since confocal laser scanners (CLSM) would typically need a few seconds to form an image. In 1991, the first in viva images at the surface and below the surface of the skin were obtained by using a TSM-based ophthalmoscope (6). From this first attempt, a commercial TSM was redesigned in 1993 (7) for in viva skin imaging with the development of a skin-contact stabilization device fitted around an objective lens that flattened the skin and limited image shifts. The Anderson team (8) solved the slow scan drawback of the CLSM in 1995 by introducing a rotating polygon mirror in the scanning path that produced video-rate images of the human skin in viva. We describe in this report the design of a new video-rate CLSM prototype adapted to the in viva exploration of human skin that produces sharper reflected-light images. It has been used to depict the various epidermal layers. An explanation of the strong reflection of the basal keratinocytes is proposed. The influence of sun exposure on cellular damage and on melanin migration was likewise investigated. MATERIALS AND METHODS THE IN VIVO CLSM PROTOTYPE The prototype was constructed around the compact Oz fast-scanning confocal module manufactured by Noran Instruments Inc. and with the contact-stabilization device designed in-house. A Nikon x40, 1.30 N.A., oil-immersion objective lens of 160-mm tube length was screwed into a microscope tube directly attached to the scanning-light exit of the confocal module. The rapid Z direction scan (up to 30 optical sections per second) was performed by moving the objective lens with a piezoelectric driver having a maximum travel of 350 pm. The contact device was centered in front of the objective lens. The distance between the lens and the contact tip can be adjusted with coarse and fine manual screws (Figure 1). In this configuration, it is the objective lens that moves vertically while the skin surface remains fixed. The confocal scanner has a two-channel configuration that permits simultaneous imag- ing of reflected (scattered) light and fluorescence contrast. A galvanometer-mounted mirror was used to scan the slow axis while an acousto-optic deflector (AOD) was used to scan the fast axis. The latter was a solid-state device that provided variable scanning rates ranging from several seconds per image up to 30 images per second and higher. It was designed to deliver 1000 resolved points within its operating range. The confocal head was suspended from a cantilever arm that gave access to the skin surface of the whole human body (9). Easy positioning of the scanning head was provided by a motorized stage. The light source consisted of a Krypton/Argon laser with power limited to 2 mW at the objective lens in order to avoid photodamage to the skin tissue. The laser provided three wavelengths: 488,568, and 647 nm. A Silicon Graphics workstation was used to control the system and also performed the capture of a digital image stream generated by the control electronics. The acquisition time was limited only by the