High-definition optical coherence tomography intrinsic skin ageing assessment in women: a pilot study

Several non-invasive two-dimensional techniques with different lateral resolution and measurable depth range have proved to be useful in assessing and quantifying morphological changes in skin ageing. Among these, only in vivo microscopy techniques permit histometric measurements in vivo. Qualitative and quantitative assessment of chronological (intrinsic) age-related (IAR) morphological changes of epidermis, dermo-epidermal junction (DEJ), papillary dermis (PD), papillary-reticular dermis junction and reticular dermis (RD) have been performed by high-definition optical coherence tomography in real time 3-D. HD-OCT images were taken at the internal site of the right upper arm. Qualitative HD-OCT IAR descriptors were reported at skin surface, at epidermal layer, DEJ, PD and upper RD. Quantitative evaluation of age-related compaction and backscattered intensity or brightness of different skin layers was performed by using the plugin plot z-axis profile of ImageJ® software permitting intensity assessment of HD-OCT (DICOM) images (3-D images). Analysis was in blind from all clinical information. Sixty, fair-skinned (Fitzpatrick types I–III) healthy females were analysed retrospectively in this study. The subjects belonged to three age groups: twenty in group I aged 20–39, twenty in group II aged 40–59 and twenty in group III aged 60–79. Only intrinsic ageing in women has been studied. Significant age-related qualitative and quantitative differences could be noticed. IAR changes in dermal matrix fibers morphology/organisation and in microvasculature were observed. The brightness and compaction of the different skin layers increased significantly with intrinsic skin ageing. The depth of visibility of fibers in RD increased significantly in the older age group. In conclusion, HD-OCT allows 3-D in vivo and real time qualitative and quantitative assessment of chronological (intrinsic) age-related morphological skin changes at high resolution from skin surface to a depth of the superficial reticular dermis.


Introduction
Skin ageing has become an important health market [34]. Many of the treatments offered claim to modulate processes involved in skin ageing. Testing the efficacy of these therapies is important for consumers, dermatologists, cosmetic industry and regulatory authorities [36,44].
For the assessment of molecular mechanisms involved in intrinsic skin ageing, invasive tests are clearly the gold standard [2,32,50]. However, the invasive nature of skin biopsy is not the most suitable method to investigate skin ageing in the general population. Skin ageing is a physiological process and, for obvious ethical reasons, efficacy testing of anti-ageing treatments should therefore be based on non-invasive methods [11].
High-definition OCT (HD-OCT) is a recently introduced non-invasive technology based on the principle of low coherence interferometry [4][5][6][7][8][9]15]. This method permits real time three-dimensional (3-D) imaging with cellular resolution up to 570 lm depth. This enables visualization of cells in their micro-architectural environment at up to the superficial reticular dermis. It was recently demonstrated that real time 3-D imaging provides accurate information on dermal matrix fibre organisation and microvasculature volume [9].
The aim of this study is the qualitative and quantitative 3-D HD-OCT assessment of intrinsic age-related (IAR) morphological skin changes [46] from skin surface up to the superficial reticular dermis.

Study sample
HD-OCT images of 60, fair-skinned, healthy females (skin types I-III) were retrieved from first author's private practice for inclusion in this retrospective study. Inclusion criteria were (1) availability of good quality HD-OCT images of normal skin at inner site of upper arm and (2) absence of signs of actinic damage in this anatomic region. These images were taken during daily practice as control/ reference HD-OCT images in comparison with HD-OCT images of affected skin. We conformed to the Helsinki Declaration with respect to human subjects in biomedical research. All international rules governing clinical investigation of human subjects were strictly followed. Approval from local ethical committee and informed consent from all participants were obtained. Moreover, this study affected neither the routine diagnosis nor treatment of the lesions presented by the included subjects.

Image acquisition by HD-OCT
Non-invasive 3-D imaging by HD-OCT (Skintell Ò , AgfaHealthcare, Mortsel, Belgium) has been used to image the internal site of the right upper arm by holding . Papillary rings are larger and more irregular on en face imaging (green circle). DEJ is thinner compared to young adult. g-i Older aged women: a disarranged furrow pattern is observed with some linearization of the furrows (dark green arrows). ET is thinnest in this group (red vertical lines). The DEJ is flat and very thin (green arrow). This results in almost absence of papillary rings on en face image the probe aligned with the axis of the humerus. Instruments and acquisition methods have been previously described [4][5][6][7][8][9].

Evaluation of IAR morphological HD-OCT features
All images were evaluated by first author, in blind from any clinical information.
Following Z-levels were scanned: skin surface, dermoepidermal junction (DEJ), upper papillary dermis (up-PD), lower papillary dermis (low-PD) and upper reticular dermis (up-RD). The junction between PD and RD is represented by the highest peak after the valley [33].

Qualitative evaluation
Standard colour setting (Table 1; Fig. 1) 1. Furrow patterns have been evaluated as previously described [51]: small rhomboidal, large rhomboidal, linear or disarranged at skin surface on en face images, 2. Flattening (effacement) of DEJ on cross-sectional images and altered ringed pattern of papillary rings on en face images [30] Inverted colour setting: fibers assessment (

Quantitative evaluation
Backscattered intensity assessment of 3-D HD-OCT DICOM (digital imaging and communication in medicine) images was achieved by using the plugin plot z-axis profile of ImageJ Ò software. This is an open source image processing program designed for scientific multidimensional images. The procedure is explained in Fig. 5. The quantitative evaluation included compaction (measured on the x-axis: from 0 to 570 lm axial position or depth) and evaluation of backscattered intensity ''brightness'' (measured on the y-axis from 0 to 800 arbitral units (AU). (see Fig. 6 for details). In addition the depth of visibility of fibers in RD was measured. The junction between PD and RD is represented by the highest peak after the valley [33].

Statistical analysis
One-way analysis of variance (ANOVA) was used to compare means of three samples using the F distribution. Moreover, Scheffé test was used for all pairwise comparisons. Calculations were made by using MedCalc Ò statistical software version 14.12.0.
All dichotomous variables describing the presence/absence of particular HD-OCT features of skin ageing were included. Absolute and relative frequencies were calculated for different age groups. Chi-squared (v 2 ) test was employed to compare each age group versus the other age groups. The phi (u) coefficient, employed to weight diagnostic power of each significant parameter, is a measure of association of two binary variables and is related to the Chi-squared (v 2 ) statistic by the formula: u 2 = v 2 /n, where n equals the total number of observations.

Discussion
In this study we presented morphological features of sunprotected skin visualized in 3-D by HD-OCT in women belonging to three different age groups. The terminology and study design of the in vivo confocal microscopy studies [30,51] dealing with skin c Fig. 3 HD-OCT dermal ultrastructural and microvascular features at the junction of papillary and reticular dermis in women of three different age groups. Young women: a cross-sectional image with dark blue horizontal line indicating the Z value of the en face image. Short black line with Z value just under basal cell layer. The thickness of the PD is shown by a double yellow arrow. The sub-papillary plexus corresponds with larger, elongated and horizontal oriented hyporeflective spaces (green arrows). b Corresponding en face image displays a randomly oriented network of coarse, long, curved but discrete rope-like bundles of fibers (green circle). Disperse black dots align with these fibers (light blue arrow). Clusters of black dots are sparse. The sub-papillary plexus corresponds with larger, elongated and horizontal oriented hyporeflective spaces. (green arrows). Middle aged women: c cross-sectional image with dark blue horizontal line indicating the Z value of the en face image. Short black line with Z value just under the basal cell layer. The thickness of the PD is indicated by a double yellow arrow. The papillary thickness is smaller compared to young adults. d Corresponding en face image. Fibers are thicker, longer and straighter compared to young adults. They form more marked rope-like bundles of fibers (green circle). Moreover, they become more aggregated and aligned in few directions corresponding with furrow pattern (dark green lines). Disperse black dots condensed but still aligned with fibers (light blue arrows). Clusters of black dots increase in density and are progressively oriented in a more horizontal way (light blue arrow). The subpapillary plexus corresponds with larger, elongated and horizontal oriented hyporeflective spaces. (green arrows). Older aged women: e cross-sectional image with dark blue horizontal line indicating the Z value of the en face image. Short black line with Z value just under the basal cell layer. The thickness of the PD is indicated by a double yellow arrow. The papillary thickness is smallest. f Corresponding en face image. Straight rope-like bundles of fibers (green circle) aggregate and align in one main direction corresponding with the furrow pattern (dark green lines). The Clusters of black dots merge to form compact dark blotches. The sub-papillary plexus becomes very prominent with strongly dilated horizontal oriented elongated branching hyporeflective spaces (green arrows) ageing related morphological changes of epidermis and superficial dermis have been adapted to the HD-OCT. A new terminology and adjusted study design have been proposed regarding morphological assessment of fibers in both papillary and superficial reticular dermis as well as dermal microvasculature in inverted colour setting. Moreover, the present study offers for the first time a quantitative evaluation of HD-OCT descriptors for intrinsic skin ageing based on backscattered intensity measurements.
HD-OCT enables imaging of IAR qualitative skin changes. HD-OCT permits the visualization of the surface texture and furrow pattern in one single en face image because of the large field of view (1.8 9 1.5 mm). This paper suggests that loss of intersecting furrows in sunprotected skin sites is a chronological process, not necessarily linked to sun damage. These findings are in line with previous studies regarding RCM and skin surface topography [1,30,51].
With intrinsic ageing, the flattening of the DEJ on crosssectional HD-OCT imaging is more pronounced. The higher axial resolution (3 lm) of HD-OCT probably enables a better visualization of the DEJ compared to other non-invasive technologies with cross-sectional imaging such as HF-US and conventional OCT. According to Lavker et al. the major change in ageing skin is the flattening of the DEJ because of retraction of the epidermal down-growths in combination with a loss in proliferative capacity associated with the aged epidermis [26]. Age-related functional and structural changes in human DEJ components have been described [27]. The flattening of the DEJ on HD-OCT cross-sectional images corresponds with irregular papillary rings up to the complete disappearance of these rings on en face images. These findings are in line with previous RCM observations by Longo and Wurm et al. [30,51].
However, age-related morphological changes of the dermal matrix fibers could be observed with HD-OCT. Moreover, real time 3-D HD-OCT provided volumetric information about the dermal matrix fibers organisation. With intrinsic skin ageing, fibers in PD become thicker, longer, straighter and progressively aggregated in a lace-like network. In the superficial RD these fibers formed IAR straight thick ropelike bundles, no longer oriented randomly but in only in few directions. These findings are in line with evidence for the IAR degradation of fibrous extracellular matrix components including elastin, fibrillin-containing oxytalan fibers and the collagen types I, III and IV [32]. Interestingly only in OA-skin very long ([400 lm), almost unidirectional fibers were imaged by HD-OCT at up to a 185 lm depth under the PD/RD junction. Moreover, these fibers were aligned with the linear furrow pattern. This phenomenon probably corresponds to the process of glycation producing crosslinks between macromolecules, which provides an explanation for the increased age-related stiffness of the skin [35]. In PD, dispersed dark dots (imaged in the inverted colour setting mode) aligned with fibers and vessels and become more and more condensed with intrinsic ageing. In young skin, these dark dots had a candlestick-holder-like 3-D configuration as if they were sustaining the dermal papillae. In elderly skin these black dots condensed progressively to form large dark blotches near the flattened DEJ. To the best of our knowledge, these observations have not been described in studies using other in vivo microscopy techniques. Our findings seem to be in agreement with observations made by scanning electron micrography of matrix fibers in young and aged dermis (low resolution mode) [26]. The present study, however, suggests that flattening of the DEJ is related to the disappearance of the support of dermal papillae by the candlestick-holder-like configuration. IAR morphological changes of the cutaneous microcirculation were observed by HD-OCT. An age-related decrease in number and size of capillary loops in the dermal papillae and an increase in size of the sub-papillary plexus are in line with laser Doppler flowmetry and videocapillaroscopy findings [22,28,29,47]. Interestingly our study described a hyper-reflective cuff around the branched vessels of the sub-papillary plexus in elderly skin.
A quantitative assessment of HD-OCT descriptors for intrinsic skin ageing has been made possible by ImageJ Ò software analysis of HD-OCT images. A significant progressive compaction of epidermis, DEJ and PD with age could be detected. This compaction paralleled the increase in backscattered intensity or brightness of the different layers.
An IAR significant progressive compaction of the epidermal component could be found. In the present study, a new method for epidermal thickness (ET) assessment has been described. In this method stratum corneum thickness was included in the ET measurement. ET measurement by HD-OCT has already been discussed in detail by a recent investigation [8]. In that study, the thickness of the stratum corneum was not included in the overall ET assessment of the skin of the back; moreover, 55 % of the subjects belonged to YA-group and 45 % to MA-group. A mean value for ET at the back was 47.38 lm (±1.07 lm 95 % CI). In the present study, the mean value for ET at the inner site of the upper arm of subjects aged between 20 and 60 years was 54 lm (±3.75 lm 95 % CI). The difference between the means of ET in the two studies could be explained by stratum corneum in-/exclusion and anatomic site. The IAR compaction of the epidermis is in agreement  with findings by other non-invasive technologies such as RCM [30,51], MPT [23] and conventional OCT [14]. Interestingly, in the present study a significant increase in epidermal backscattered intensity (brightness) in elderly skin was observed compared to MA-and YA-skin. Skin dryness represents an important characteristic of aged skin. Aquaporin-3 distribution in human epidermis is consistent with epidermal water distribution and parallels the steep water gradient at the junction between stratum granulosum and stratum corneum [3]. A significant decrease of aquaporin 3 (AQP-3) expression in the epidermis with chronological ageing has been described [45] probably explaining the significant increase in epidermal brightness of HD-OCT images with age. A significant increase in compaction and brightness of the DEJ with chronological age was noticed with HD-OCT. This is in agreement with the IAR degradation of fibrous extracellular matrix components and with the loss of the oligosaccharide fraction which in turn impacts on the ability of tissue to retain bound water [27,32]. A decreased DEJ thickness with age has already been described for conventional OCT; moreover, there is evidence that the DEJ thickness is higher in African Americans than in Caucasians [39].
With intrinsic ageing, a more compact pattern of the fibrous dermal component of PD could be quantified and correlated with decrease in the voids or areas between the fibers of PD on HD-OCT. These areas correspond most probably to the presence of the ground substance consisting in particular of hyaluronic acid and chondroitin sulphate. The IAR loss of the oligosaccharide fraction impacts on the capacity of the PD to retain bound water [32]. On HD-OCT this loss of bound water resulted in higher brightness. The IAR compaction of the dermis is in agreement with other non-invasive technologies such as MPT [23], conventional OCT [33] and HF-US [16-20, 41, 43].
The highest peak after the valley (Fig. 6) corresponds with the junction between PD and RD as described for conventional OCT [33]. Moreover, the depth of visibility of fibers in reticular dermis increased dramatically in the OA group. This could be explained by fibre rearrangements and alterations such and glycation of collagen fibers [32,35].
This pilot study has some limitations: (1) only IAR changes in skin morphology have been studied, with no comparison with sun-exposed areas; (2) only women were assessed; (3) individuals with significant systemic comorbidities were excluded and last but not least (4) no histological validation of IAR qualitative skin changes has been performed.
In conclusion, HD-OCT permits to assess qualitatively and quantitatively in vivo and real time three-dimensional IAR morphological skin changes in high resolution from the skin surface to the superficial reticular dermis. This could offer a new possibility to test the efficacy of different cosmetic products. Moreover, HD-OCT assessment of these changes could provide interesting additional information regarding the biological age of the subject as defined by the Framingham CVD risk score [12]. Furthermore, skin wrinkling at the upper inner arm has been linked to health status [48] and elastin morphology in the PD has been linked to cardiovascular diseases risk [37]. These represent interesting topics for future research.