Various light therapies have been employed in clinical practice. Advances in these treatment modalities, such as various types of laser, intense pulsed light (IPL), and light emitting diodes, have led to their wide clinical applications in treating cutaneous diseases and performing cosmetic procedures.1
The results of laser treatment depend on the immediate interaction of incident light with tissues well as the subsequent tissue regeneration and remodeling. As in many areas of medicine, histological examinations are necessary to evaluate the reaction of epidermal, dermal, vascular, and appendageal components of skin after laser therapy.2 The major drawback of conventional histological examination is its invasive nature and the inability to examine the light-tissue interaction in real time. The use of inappropriate parameters can result in inadequate treatment outcome and may lead to serious side effects.3 To be specific, skin conditions can vary among individuals as well as among the treated areas in the same individual. To achieve a favorable result, the parameters employed should be adjusted accordingly. Despite the wide application of light therapy, the association between immediate light-tissue reaction, subsequent tissue remodeling, and final clinical outcome has not been investigated in detail due to the lack of an appropriate real-time noninvasive monitoring tool. As a result, validation of appropriate adjustable clinical treatment parameters often relies on repetition of trial-and-error practice.
Investigators have resorted to noninvasive objective methods to assess laser-tissue reaction, such as sonography and optical coherence tomography.4, 5 However, limited by the resolution, these techniques do not enable detailed analysis of cutaneous structural alterations. Recently, reflectant confocal microscopy has gained popularity as a useful clinical tool for noninvasive imaging by providing subcellular resolution.6 Though it is powerful in analyzing the cellular morphologies in epidermis, its ability to resolve the turbid dermal extracellular matrical structures is limited.
Recently, the nonlinear optical imaging technique of multiphoton microscopy (MPM) has emerged as a useful technique for skin imaging.7, 8, 9, 10 In MPM, fluorophores are excited by simultaneous absorption of two or more photons in the near-IR spectrum as opposed to absorption of one photon in the UV or visible light spectrum in conventional florescence microscopy. To achieve efficient multiphoton excitation, extremely high peak photon input from lasers with pulses in the range of tens to hundreds of femtoseconds is required.7 Due to the nonlinearity, the excitation occurs only at the focal point and images from a thin optical section can be acquired without using a pinhole. MPM holds several unique advantages over confocal scanning laser microscopy, such as reduced specimen photobleaching, higher contrast images, and enhanced penetration depth.7, 9, 11 Since endogenous autofluorescent substances including NAD(P)H, keratin, and elastic fibers can be excited by multiphoton excitation, label-free imaging can be achieved.9, 11
By use of a femtosecond laser, the nonlinear polarization effect of second-harmonic generation (SHG) can also be used for visualizing biological structures lacking inversion symmetry.11, 12 For SHG, the interaction of noncentrosymmetric structures with incident light yields photons with a wavelength of exactly half the wavelength of the incident light. Unlike fluorescence excitation, the polarization effect of SHG does not involve absorption of the incident photons by the molecules. Therefore, photobleaching and thermal damage can be avoided. Collagen fibers, myosin, and microtubules are effective biological SHG signal generators.12 As SHG is highly sensitive to structures, SHG signals can be used for probing modification of collagen structures in cancer and evaluating the thermal structural transitions of collagen.13, 14 The wavelength of SHG is shorter than the fluorescence generated by multiphoton excitation, and hence, SHG signals can be easily separated from fluorescence for further analysis and image processing. The combination of multiphoton autofluorescence (AF) and SHG imaging has been applied to evaluating cutaneous aging and differentiating basal cell carcinoma from normal dermal stroma.15, 16 Multiphoton microscopy has been used to detect corneal architecture and intratissue femtosecond laser effect in a rabbit eye model.17
In this paper, we assessed the usefulness of MPM in monitoring laser-skin reactions in vivo. An erbium:YAG (Er:YAG) laser was chosen in this study due to its wide dermatological applications, such as dermabrasion, skin tumor removal, skin rejuvenation, and enhancement of transcutaneous drug delivery.18, 19 We investigated the morphological and quantitative changes of AF and SHG signals associated with Er:YAG laser irradiation and evaluated the potential of MPM in real-time evaluation of laser-skin interaction.
Materials and Methods
The multiphoton microscope is similar to the one we previously described with modification.20 A femtosecond titanium-sapphire (ti-sa) laser (Tsunami; Spectra Physics, Mountain View, California) pumped by a diode-pumped solid state (DPSS) laser (Millennia X; Spectra Physics, Mountain View, California) was used as the excitation source. The output of the laser system was guided toward a modified commercial upright microscope (E800; Nikon, Tokyo, Japan). Prior to entering the microscope, the excitation source was angularly deflected by an scanning system (Model 6220; Cambridge Technology, Cambridge, Massachusetts). The input of the upright microscope was modified to accommodate a beam expander. The excitation source was beam expanded and reflected toward the focusing objective [S Fluor , numerical aperture ; Nikon, Tokyo, Japan] by a short-pass primary dichroic (700DCSPXRUV-3p; Chroma Technology, Rockingham, Vermont). The power at the sample was . The nonlinear AF and SHG signals were generated at the sample focal plane and collected in the epi-illuminated or back-scattering geometry by the same focusing objective. After passing through the primary dichroic, the AF and SHG signals were separated into two separate channels where they were detected by independent photomultiplier tubes (R7400P; Hamamatsu, Hamamatsu City, Japan). The AF and SHG signals were separated by a secondary dichroic (435DCXR; Chroma Technology Rockingham, Vermont). The secondary dichroic is a long-pass filter that reflects below and transmits above . The SHG signal was further filtered by a SHG filter (HQ390/20; Chroma Technology, Rockingham, Vermont) centered at with a bandwidth of , while the AF was detected by a bandpass filter (E700sp-2P-E435lp; ChromaTechnology, Rockingham, Vermont) for broadband fluorescence detection between 435 and . To acquire a global image of the thermally treated specimens at high resolution, an sample positioning stage (H101; Prior Scientific Instruments, Cambridge, UK) was used to translate the skin specimen after each imaged frame. An array of overlapping and individually beam-scanned images, each with an area of , was then assembled.
Image Processing and Analysis
The MPM images were analyzed and reconstructed by the software ImageJ 1.36b (National Institutes of Health, Bethesda, Maryland). The toolbox enabled visual inspection of various morphologic features and fluorescence patterns of cellular components within normal skin and laser-irradiated skin. AF and SHG signals of each site were quantitatively acquired. To determine whether treatment with Er:YAG laser at various fluences influences dermal SHG, average SHG intensity at a depth of more than of untreated and treated skin were analyzed with a general linear model. A value of , corresponding to the 95% confidence interval, was selected for analysis of significant difference of SHG in this study. SPSS 15.0 statistical software (SPSS Inc., Chicago, Illinois) was used.
Er:YAG Laser Assembly
An Er:YAG laser (Burane; WaveLight Laser Technologie AG, Erlangen, Germany) was used to irradiate skin. The laser had a wavelength of and pulse duration of . An articulated arm was used to deliver the laser beam onto the skin. Laser beam spot size was in diameter. Each skin site was exposed once.
Accommodation of the Mouse on the Microscopic Stage
The experimental protocol involving animals was approved by our Institutional Animal Care and Use Committee. Male nude mice (Balb/c-nu strain, old) were anesthetized by intraperitoneal phenobarbitol injection and prepared for Er:YAG laser irradiation. Immediately after laser irradiation, mouse dorsal skin were stabilized in a dorsal skin fold chamber, coverslipped, and observed under the MPM.
Multiphoton Image of Unirradiated Nude Mouse Skin
Figure 1 shows the multiphoton images of unirradiated nude mouse skin. MPM images of skin surface reveal compact anucleated corneocytes. Autofluorescent pentagonal or hexagonal corneocytes are tightly packed and the cell borders are clearly delineated [Figs. 1a and 1b]. Hair is also intensely autofluorescent and can be seen on the skin surface (arrows). In the optical sections corresponding to the stratum spinosum and stratum basale, the cells are characterized by fluorescent cellular cytoplasm surrounding nonfluorescent round nuclei. A gradual decrease in cell size can be observed when serial images are taken from the stratum spinosum down to the stratum basale. The ratio of the nucleus to the cytoplasm also increases with depth [Figs. 1c and 1d]. The dermal-epidermal junction with intermingled AF signals from basal keratinocytes and SHG signals from collagen fibers can be seen in Fig. 1d. Collagen fibers in dermis emit strong SHG signals [Fig. 1e].
In the corresponding histological image, in contrast to compact structures revealed in the MPM images, stratum corneum is loose with a basket-weave pattern [Fig. 1f]. This discrepancy is due to the artifacts caused by fixation and staining procedures in histological examinations that disrupt the normal intercellular packing of corneocytes.21, 22, 23 In addition, the epidermis is thinner measured histologically (about ) than that measured by multiphoton microscopy (about ). This indicates that tissue shrinkage during processing of the specimen for histological examination can be a significant artifact that interferes with the precise measurement of tissue thickness.
Selective Disruption of Stratum Corneum by Er:YAG Laser at a Low Fluence of Revealed by Multiphoton Imaging
Nude mouse skin was irradiated with Er:YAG laser at a fluence of . In the MPM images, compared with the tightly packed hexagonal corneocytes with distinctive intercellular borders in control group [Fig. 1a], the morphology of a large proportion of corneocytes has been disrupted into a homogenized pattern and intercellular borders can not be clearly delineated [Figs. 2a and 2b ]. In addition to the homogenization change, selective loosening of stratum corneum and loss of corneocytes can also be observed focally [Fig. 2a, asteroids]. Brightly fluorescent hairs are still intact and are relatively resistant to laser irradiation (arrows). However, in the corresponding histological image, only scanty corneocytes are still preserved [Fig. 2f]. We speculate that the homogenized or loosened corneocytes are lost during histological processing.
When images are taken farther down the skin surface, the morphology of keratinocytes in the stratum spinosum and stratum basale is still intact [Figs. 2c and 2d]. The SHG signals can still be detected from the dermal collagen [Fig. 2e]. In the histological image, epidermal keratinocytes are also still intact and no significant change of dermal collagen is revealed [Fig. 2f]. Since the AF from the epidermal cells and SHG from the dermis are all decreased and the cellular morphology and collagen fibers are relatively intact, the decrease of the intensity of AF and SHG from epidermal cells and dermal collagen may be caused by the shielding effect from the intensely fluorescent disrupted corneocytes.
Ablation of Stratum Corneum and Thermal Injury to Underlying Structures by Er:YAG Laser at Moderate Fluence of
After irradiation at the fluence of , partial ablation of stratum corneum is revealed and the residual stratum corneum displays a loose reticular pattern [Figs. 3a and 3b ]. The stratum spinosum starts to become a homogenous fluorescent morphology without discernable intercellular borders [Fig. 3c]. Residual discernable keratinocytes are mixed with homogenized cells in basal layer [Fig. 3d]. In the corresponding histological image, intact corneocytes can hardly be seen. There is a thin layer of eosinophilic debris in which upper epidermal cells show a condensed nucleus and more eosinophilic cytoplasm. Some of these cells are partially detached from the underlying epidermal cells. Many residual basal cells become vacuolated [Fig. 3f, arrowhead].
When the images are taken farther down to the dermis, the intensity of SHG signals from the dermis is greatly diminished and the fibrous morphology of collagen is further deteriorated into an amorphous structure due to the residual thermal damage from laser irradiation [Fig. 3e]. Compared with MPM images, this denatured collagen zone can not be easily appreciated in the histology. It presents as a thin layer of slightly homogenized collagen immediately beneath dermal-epidermal junction [Fig. 3f].
Complete Ablation of Epidermis with a Thermal Injury Zone to the Dermis by Er:YAG Laser at a High Fluence of
High-fluence laser irradiation produced a depressed crateriform hole [Fig. 4a ]. On the rim of the crateriform hole, the residual corneocytes and keratinocytes are homogenized into a highly fluorescent ring [Fig. 4a]. In the central area of the crateriform hole, the stratum corneum can not be seen after being ablated. When images are taken down to the skin in the central area [Figs. 4b, 4c, 4d, 4e], intact keratinocytes can not be observed. In the dermis, no SHG signals can be detected [Fig. 4e]. In the histological image, no residual epidermal cells are observed [Fig. 4], and the crater can not be easily appreciated due to tissue shrinkage and deformation. Eosinophilic homogenized degeneration of dermal collagen can be seen in the central area of the ablated zone [boxed area, Fig. 4f].
Probing the Laser-Induced 3-D Structural Alterations by Multiphoton Imaging
To highlight the spatial impact of laser irradiation on skin, the serial multiphoton images are reconstructed into 3-D images (Fig. 5 ). In unirradiated skin, keratin and keratinocytes form an autofluorescent layer from the top of the image and fibrous collagen emitting SHG signals are penetrating in the dermis [Fig. 5a]. The dermal-epidermal junction shows an undulating plane between the AF and SHG signals. The crater generated by high-fluence Er:YAG laser irradiation can be more easily visualized in the reconstructed image [Fig. 5b]. The shape of the crater is consistent with the reverse bell-shaped Gaussian distribution of the laser beam. The absence of SHG signal in residual dermis indicates that the collagen is denatured in the residual thermal zone.
Quantitative SHG Analysis Detects the Thermal Injury Profile in Laser-Irradiated Skin
It has been reported that the change in the intensity of SHG is a good predictor of collagen structural disruption.13, 14, 24 We quantitatively analyzed SHG intensity after laser irradiation at various fluences. Figure 6 shows the SHG intensity profiles of normal skin and skin irradiated with a laser at various fluences. In the unirradiated skin, the intensity of SHG started to increase between 40 and from the surface. This zone corresponds to the undulating dermal-epidermal junction. SHG intensity approaches its plateau at the depth of about from the surface. This corresponds to the plentiful collagen content in the dermis. After irradiated with Er:YAG laser at low fluence of , the plateau intensity of SHG signals in the dermis decreases compared with the control group. However, the increase of SHG intensity at the dermal-epidermal junction of depth about is still observed. Since the fibrous structures of collagen are still observed, the decrease in the plateau SHG intensity in dermis may be caused by the shielding effect of the overlying homogenized highly fluorescent stratum corneum. At the moderate fluence of , SHG decreases more significantly and the increasing trend of intensity at the dermal-epidermal junction can only be barely observed. In skin exposed to a high fluence of , the SHG signals are undetectable. SHG signals of dermal collagen are significantly different between four groups with paired comparisons .
The Er:YAG laser and the laser are two commonly used resurfacing lasers. The epidermis is composed of 80% water. In these two lasers, the major chromophore for absorbance is water. At high fluences of the laser for ablation, instant heating of water to occurs and vaporization of tissue ensues. However, it can also yields a zone of coagulation beyond the vaporization zone to a variable degree, depending on the mode and parameters of laser used.25 The residual thermal effects in the temperature range of can result in denaturation of proteins, unwinding of DNA, and disruption of cell membranes. Though the residual thermal zone of the laser can contribute to tissue shrinkage for the purpose of tightening in skin rejuvenation, it can also lead to undesirable effects. For example, this residual thermal effect can compromise cell function and necrosis ensues.26
With the Er:YAG laser at , the energy is so efficiently absorbed by water (12 to 18 times that of the wavelength) that its shorter penetration depth results in a more precise zone of tissue ablation with reduced residual thermal injury compared with laser.27 The Er:YAG laser is currently widely used in dermatology as an ablative tool for skin tumor removal, acne scar treatment, and facial rejuvenation. The vaporizing effect of Er:YAG laser ablates superficial tissue, whereas the residual heat conduction can still lead to tissue coagulation and collagen denaturation.2 Though the residual thermal zone can induce collagen shrinkage and neogenesis,28 too much thermal damage can also lead to unwanted complications, including postinflammatory hyperpigmentation and scars.3 The evaluation of residual thermal zones previously depended on histological examinations, but the procedure does not allow for real-time imaging. When the collagen is totally coagulated, eosinophilic homogenization can be seen in histological examination. However, if collagen is not heated enough to become totally coagulated, it can be difficult to detect the subtle changes in histology.24, 29 Thermal denaturation results in the loss of molecular order in the collagen fibers, which can have a profound influence on SHG.13, 14, 24 Consistent with our previous reports,24, 29 thermal treatment can disrupt regular collagen packing and hence the SHG signal in dermis decreases. Decrease in intensity of the SHG signals from collagen, which have statistically significant difference between each group, is correlated to the fluence of the laser (Fig. 6). The findings in this study support that fact that the SHG signal in MPM is a more sensitive tool in identifying the degree and extent of residual thermal damage in laser-irradiated skin. Hence, the SHG signals may be used as a real-time feedback system during laser treatment to visualize and quantify the residual thermal zone.
In addition to tissue ablation by Er:YAG laser at high fluence, Er:YAG laser at low fluence can enhance the transdermal delivery of many lipophilic and hydrophilic drugs and materials, including 5-aminolevulinic acid, 5-fluorouracil, vitamin C, and even macromolecular DNA.19, 30, 31, 32 Though the detailed mechanism is unknown, it is believed that stratum corneum might be disrupted or partially ablated to achieve such a penetration enhancing effect. Previously, we could only assess the stratum corneum structural alterations in histology. However, the processing for histological examination compromises the compact structures of stratum corneum and can lead to detachment of superficial corneocytes [Fig. 2f]. Another drawback of histology is that thin vertical sections do not allow large-area or 3-D analysis of stratum corneum structures. The advantages of MPM in imaging stratum corneum over conventional histological examinations include the abilities to image skin in vivo without processing artifact and to provide more information by en face large-area and 3-D imaging.33 In this paper, we demonstrate that the structural changes of stratum corneum after low-energy Er:YAG laser treatment can be better highlighted by MPM than histological examinations. By use of MPM, we show that Er:YAG laser can selectively disrupt the compact structure of the stratum corneum at a low fluence while little ablative effect takes place (Fig. 2). The homogenized morphology of cornecytes is similar to the structural alterations induced by a depilatory cream, which has been reported to be able to enhance the transdermal penetration of both small model drugs and insulin.33, 34, 35 This is the first report of stratum corneum structural alterations after low-fluence Er:YAG laser irradiation by use of MPM. It has been demonstrated that, in combination with appropriate model fluorescent drugs, MPM can be used to study transcutaneous drug delivery in vitro.33, 36, 37, 38 Our findings suggest that MPM can be an ideal tool for studying the mechanism and dynamics of laser-assisted transdermal drug delivery in vivo.
There is a Gaussian distribution of intensity across most kinds of laser beams and this variation of irradiance across the beam can lead to variation in biological effects in the irradiated skin. The Gaussian profile has peak intensity in the center and minimal intensity at the beam edges.39 It is important to monitor the treatment effect across the laser beam to avoid over- and undertreatment in the irradiated spot. Previously, analysis of the extent and depth of tissue alterations by laser irradiation involved labor-intensive and manual-counting methods that required multiple histological tissue sections.40 In addition, the processing of skin samples for the histological examination also leads to tissue distortion and shrinkage, and it can be difficult to precisely evaluate the impact of the variation in the intensity profile across the laser beam on the laser-skin interaction. In this paper, we demonstrate that a laser-induced reverse bell-shaped slope compatible with Gaussian distribution in 3-D MPM images (Fig. 5). The laser beam has a central peak irradiance and can ablate more tissue, whereas beam edges have the lowest intensity that can only produce a shallow rim or induce nonablative thermal effects. This result can not be clearly demonstrated in the histology. Hence, MPM can be used to study the biological impact of the uneven distribution of energy within a laser beam and to facilitate the design of laser beam intensity profile to obtain a better clinical result.
In addition to ablative lasers, other lasers have been developed for specific clinical applications including treatment of vascular abnormalities, depilation, etc. Suggested treatment parameters are often based on the results from many modeling studies, which are designed to predict photothermal effect for selective photocoagulation.41 However, current modeling methods can not adequately incorporate the complex dynamical interaction between light and tissues. MPM may also be used for monitoring the biological effects of theses lasers in vivo as well. For example, with appropriate labeling, MPM enables detailed 3-D visualization of vascular structures and quantification of hemodynamics in vivo.42, 43 Cutaneous blood vessels have complex 3-D arrangement that is difficult to be appreciated in standard 2-D display in histological examination. With the reconstructed 3-D images from MPM and image analysis, it will be able to reveal the detailed vascular structures and associated changes after laser treatment and also the remodeling process.
In conclusion, by use of Er:YAG laser as the irradiation source, we demonstrate that MPM is able to probe the laser-skin reaction in vivo without labeling procedures. The selective disruption of stratum corneum at low fluence and the ablative effects as well as the residual thermal injury at higher fluences can be visualized and quantified with subcellular resolution. MPM has been used in imaging human epidermal and dermal structures in vivo.44, 45 Our results suggest that MPM can be further developed into a real-time feedback system for evaluating human responses to laser treatment in vivo.
This study was supported by grants from Far Eastern Memorial Hospital, Taiwan (FEMH-B95-01, FEMH-96-C29, and 97-FTN15), National Taiwan University Hospital (97S-842 and 97-FTN15), and National Research Program for Genomic Medicine of the National Science Council, Taiwan (NSC-96-3112-B-002-025, NSC-97-3112-B-002-037). The experiments and data analysis involving multiphoton microscopy were performed at the Optical Molecular Imaging Microscopy Core Facility, National Research Program for Genomic Medicine, Taiwan.