1.

Introduction

Port wine stain (PWS) is a congenital disease with 3 to 5‰ incidence.^{1, 2} The histopathological manifestation is the capillary dilation and malformation in the upper dermis (papillary layer). Photodynamic therapy (PDT), which has been used to treat PWS for the last 2 decade, has advantages including excellent tissue selection, strong lesion effects on microvessels, and few side effects. Therefore, it has become^{3, 4, 5, 6, 7, 8, 9, 10} one of the most effective therapies for PWS. The mechanism of PDT is that, when the endogenous or exogenous photosensitizer in the tissue is exposed to light of a specific wavelength, it absorbs the photon energy and transits to the excited state, then it soon releases energy and relaxes to its ground singlet state through a physical or chemical deexcitation process, during which a large quantity of reactive oxygen is generated. The singlet oxygen reacts with various biomolecules, resulting in cell killing through apoptosis and/or necrosis, as well as an occlusion effect on blood vessels. The key of applying PDT to PWS is to selectively destroy dilated and malformed capillaries in the upper dermis, while sparing the epidermis and deep dermis. PDT has been used for cancer treatment. Khan,¹¹ Wilson,¹² Dougherty,^{13, 14} Henderson and Dougherty,¹⁵ and Dougherty^{16, 17} used HpD as a photosensitizer combined with red light illumination and achieved good results in treating a dozen of cancer types including breast cancer, uterine cancer, basal cell cancer, and squamous cell carcinoma. The clinical applications become broader as the research on PDT goes deeper. Gu applied PDT in clinical treatment of PWS in 1991 with good clinical effects.¹⁸

Despite the fact that the effects of PDT on PWS microvessels have been widely reported, the area of vessel lesions and the relationship between treatment and efficacy are not clear due to the complex relationship among light, photosensitizer, oxygen, and tissue structure during PDT. The treatment effects in the tissue vary dynamically during the PDT process, and thus are difficult to evaluate in real time accurately. Therefore, the key issue to improve PDT therapeutical efficacy is to have an objective diagnosis and determine the proper drug and laser dosage for an individual patient. Since PWS usually occurs on the face, it is not realistic to conduct common histological observation before and after treatment. Novel noninvasive imaging methods have long being explored. The emergence of optical coherence tomography (OCT), with its main characteristics of in situ, in vivo, and in real-time use has made the noninvasive detection of skin disease possible. OCT currently has broad applications in highly scattering tissues such as the skin, blood vessels, the urinary bladder, and the splanchnic organs, as well as successful ocular applications.^{19, 20, 21, 22, 23}

Doppler OCT (DOCT), spectroscopic OCT (SOCT), and polarization-sensitive OCT (PS-OCT) have potentially powerful applications in assessing the efficacy of PDT treatment. Blood flow in PWS was measured with DOCT by Zhao ²¹ Blood oxygen saturation is an important parameter in PDT. It was measured²² by SOCT according to the absorption coefficient of hemoglobin (Hb) and oxyhemoglobin $\left(\mathrm{Hb}{\mathrm{O}}_2\right)$ at $800\mathrm{nm}$ . In addition, the depth of skin tissue lesion can be measured by PS-OCT due to birefringence of collagen in dermis.²³ Here, we apply OCT to PWS clinical detection, and obtain the images of epidermis thickness and the configuration of dilated capillary vessels. Doing this will improve the pathological diagnosis and potentially can be used to monitor PDT treatment efficacy in real time.

2.

Materials and Methods

Dilated papillaries in the PWS patient dermis is usually within $1\mathrm{mm}$ below the skin surface with an average diameter over $50\mu \mathrm{m}$ , in contrast to a $10\text{-}\mu \mathrm{m}$ diameter in normal skin vessel.^{2, 24, 25, 26, 27, 28} Therefore, we designed our OCT system with $10\text{-}\mu \mathrm{m}$ axial resolution in skin tissue, determined by light source spectral width.

OCT uses a low-coherence IR light source to detect reflection and scattering echoes from the specimen and builds images of the inner structure. Currently the time domain OCT technique is widely used. We designed and developed a portable time domain OCT system for clinical PWS detection.

Figure 1 shows a schematic of our OCT imaging system. It employs a superluminescent LED (SLED) broadband light source with a central wavelength at $1310\mathrm{nm}$ and a spectral FWHM of $70\mathrm{nm}$ . To eliminate the system reflection on the light source, the source output is first connected to an optical isolator via a pigtail fiber, and then coupled to a 92:8 fiber coupler by a three-port optical circulator. The coupler passes 92% of the light to the sample arm and 8% to the reference arm. Light reflected from the sample arm and the reference arm interferes at the coupler. It is further converted into the current signal by a- $p\text{-}i\text{-}n$ photodiode and preamplified by a current amplifier, which converts the weak current signal into a voltage signal. Before being sent to a data acquisition card, the output signal is amplified by a low-noise preamplifier and filtered by a bandpass filter. A $12\text{-bit}$ analog-to-digital (A/D) converter samples and quantifies the output analog signal. Then digital bandpass filter further removes the noise. After digital signal processing (DSP), structure OCT image is generated based on the amplitude of axial scanning. The sample arm and the reference arm are equipped with polarization controllers (PC1 and PC2) to control the polarization mode of light output.

Fig. 1

Schematic of the OCT system.

The reference arm employs the grating rapid scanning optical delay (RSOD) line, shown by the dashed-line part in Fig. 1, which realizes axial scanning with a resonant scanning frequency²⁹ at $800\mathrm{Hz}$ , carrier wave central frequency at $1163\mathrm{kHz}$ , and a $383\text{-}\mathrm{kHz}$ bandwidth. The delay line is used for the carrier wave central frequency setup and dispersion compensation, shown by the dashed frame. The grating and the swaying mirror are placed on the front and back focus planes, respectively, which can optimize dispersion match.^{30, 31}

The sample arm is equipped with a handheld detector that scans laterally. It consists of a fiber collimator, a scanning mirror and an achromatic lens. The spot size of the beam focused on the specimen determines the lateral resolution of the OCT system. The numerical aperture of the objective determines the speckle size. The detector consists of a swaying mirror and an achromatic lens. The rotation of the swaying mirror produces lateral scanning at various positions. A lens with a focal length of $20\mathrm{mm}$ is chosen, via which the beam is focused onto the sample with a focus $9\text{-}\mu \mathrm{m}$ -diam which is the lateral resolution of the OCT system. This matches the axial resolution, and therefore avoids image distortion.

In addition, all fiber ends in the system are coated with antireflection film to greatly improve the SNR. The measured SNR of the OCT system is $108\mathrm{dB}$ .

3.

Results and Discussion

The human skin is composed of epidermis and dermis. The thickness range of the epidermis is from $0.04\text{to}1.6\mathrm{mm}$ , and about $0.1\mathrm{mm}$ on average.^{27, 28} It varies greatly among individuals. The dermis is beneath the epidermis and is generally divided into two layers: the papillary layer and the reticular layer. The consideration of PDT for PWS treatment focuses on the thickness of the epidermis and the vascular distribution in the dermis papillary layer.

Forty-one Chinese patients with PWS on the face and neck (18 males and 23 females, aged from $0.9\text{to}45\text{years}$ , $8.81\pm 8.72\text{years}$ old) were recruited from among the outpatients of the Department of Laser Medicine, Chinese PLA General Hospital. This study was approved by Ethics Committee of the Chinese PLA General Hospital. OCT images of PWS skin and their contralateral normal skin were acquired from multiple sites of the 41 patients. The structural parameters, including epidermal thickness and diameter and depth of dilated blood vessels, were obtained. The average and spread of the key parameters are displayed in Table 1 . Photos and OCT images of five typical patients are displayed in Figs. 2, 3, 4, 5, 6, 7, 8, 9, 10, 11 as follows. Figures 2, 4, 6, 8, 10 show PWS areas and contralateral normal face skin, scanned by the OCT system. The scanning area is $1.8\mathrm{mm}$ (lateral) by $2.5\mathrm{mm}$ (axial) with a scanning speed of $4\text{frames}∕\mathrm{s}$ . Corresponding gray-scale OCT images (Figs. 3, 5, 7, 9, 11, where E indicates the epidermis and BV represents blood vessels) show different structures between PWS areas and contralateral normal skin. We are able to distinguish the dilated capillary vessels from normal tissue, based on blood vessel diameter and configuration. There are many “black holes” with “shadow effect” from pathological skin, in contrast to the contralateral normal skin. Normally large diameter blood vessels would produce a “shadow effect” in the B-scan OCT images because the blood has slightly more absorption of light and results in the image loss beyond the point of the proximal vessel.^{35, 36, 37, 38} From these OCT images, we obtained the similar results that “black holes” are displayed in PWS areas and rarely found in contralateral normal skin. As the average diameter of contralateral vessels in normal skin is²⁶ about $10\mu \mathrm{m}$ , the vascular structure of normal skin cannot be displayed clearly with the resolution of our current OCT system. However, the diameter of dilated capillary vessels in patients with PWS are usually²⁶ larger than $50\mu \mathrm{m}$ . Therefore, the dilated capillary network in the papillary layer in PWS patients can be distinguished.

Fig. 2

Areas scanned by the OCT system of patient 1.

Fig. 3

OCT images of right lower eyelid, right upper lip, and their contralateral normal skin of patient 1.

Fig. 4

Areas scanned by the OCT system of patient 2.

Fig. 5

OCT images of left outer canthus, left pars zygomaticus, and the contralateral normal skin of patient 2.

Fig. 6

Areas scanned by the OCT system of patient 3.

Fig. 7

OCT images of right pars zygomaticus, right buccal division, and the opposite normal skin of patient 3.

Fig. 8

Areas scanned by the OCT system of patient 4.

Fig. 9

OCT images of right lower eyelid, right upper lip, and the contralateral normal skin of patient 4.

Fig. 10

Areas scanned by the OCT system of patient 5.

Fig. 11

OCT images of left anterior auricular, left coner of the mouth, and the contralateral normal skin of patient 5.

Table 1

Parameters of epidermis and dilated blood vessels determined by OCT images ( n=41 , data presented as mean±standard deviation).

The measured mean diameter and depth of the vessels from 41 PWS patients in our study are $94.61\pm 20.09\mu \mathrm{m}$ and $360.70\pm 50.20\mu \mathrm{m}$ , respectively. The mean blood vessels diameter fit well with the typical biopsy and confocal microscopy study²⁶ $(87.72\pm 3.21\mu \mathrm{m})$ . The mean blood vessels depth from OCT images correlates (although it is somewhat smaller) with those results obtained by the biopsy (Barsky ² and Zhou ²⁴ obtained the mean vessel depths of $0.46\pm 0.1$ and $0.45\pm 0.2\mathrm{mm}$ , respectively). The difference might be due to the OCT penetration limit where some blood vessels could not be detected in deeper tissue. Furthermore, thickness of epidermis was obtained in Table 1. The measured mean epidermal depth of PWS skin and contralateral normal skin are $64.61\pm 16.60$ and $64.83\pm 17.01\mu \mathrm{m}$ , respectively. There is no statistical difference between them.

PDT for PWS requires the simultaneous presence of a laser, a photosensitizer, and singlet oxygen. Therapeutic efficacy depends on laser dosage, photosensitizer concentration, and tissue oxygenation level. Optimal treatment outcome can be achieved only when these three parameters are closely matched. Epidermal thickness correlates with laser dosage, due to the epidermal absorption and scattering of light. For example, adults are given $100\mathrm{mW}∕{\mathrm{cm}}^2$ laser treatment in PDT, while $60\mathrm{mW}∕{\mathrm{cm}}^2$ laser intensity is given to children, since the adults epidermis is thicker than that of children.^{39, 40}

The diameter of blood vessels relates with the oxygenation levels in PDT. In our previous mathematical model study,⁴¹ lower oxygenation levels distribute in large-diameter blood vessels than that in small blood vessels. This is why PDT for light-red-color PWS patients usually achieves a better therapeutic effect than that for dark-red-color ones. In addition, clinicians observe that the first PDT therapy often achieves good results, although the therapeutic effect decreases with the number of therapies in the same PWS patient. We speculate that the shallow blood vessels have been first removed, therefore the same laser dosage is not enough to remove deep blood vessels. The depth of blood vessels from OCT images will be helpful to study the relationship between therapeutic effect and therapeutic times.

In the past, when applying PDT to PWS treatment, doctors chose drug dose and laser parameters (wavelength, exposure duration, laser power intensity) mainly according to the color of the red stain (the color of PWS areas mainly due to dilation degree of capillary vessels in papillary layer) and skin hyperplasia. It is difficult to diagnose pathological types because the thickness of skin varies from patient to patient and the thickness distribution of skin in different part of the same patient also varies. There are certain errors to adopting a similar treatment to different pathological types. Furthermore, diagnosis of pathological types is influenced by pigments in patient skin and by visual judgment of doctors. We are currently investigating 3-D OCT and Doppler OCT, which will provide further details on the configuration and blood flow of the capillary vessels, epidermis structures, papillary layers, and dermis. Doctors can make objective judgments of pathological types according to the thickness of the epidermis and the diameter and depth of dilated blood vessels. In conclusion, the anatomical parameters provided by OCT are very meaningful for both clinical treatment dosage and the judgment of treatment effects.

4.

Conclusions

We implemented an OCT system capable of distinguishing the dilated capillary vessels from normal tissue and the thicknesses of the epidermis and the papillary layer. The imaging rate of the OCT system is $4\text{frames}∕\mathrm{s}$ with axial and transverse resolution being 10 and $9\mu \mathrm{m}$ , respectively in the skin tissue. The imaging results demonstrated that OCT is a promising tool of noninvasive “optical sectioning” for clinical PWS treatment.