Non‐invasive optical biopsy by multiphoton microscopy identifies the live morphology of common melanocytic nevi

Abstract Multiphoton microscopy (MPM) is a promising non‐invasive imaging tool for discriminating benign nevi from melanoma. In this study, we establish a MPM morphologic catalogue of common nevi, information that will be critical in devising strategies to distinguish them from nevi that are evolving to melanoma that may present with more subtle signs of malignancy. Thirty common melanocytic nevi were imaged in vivo using MPM. Quantitative parameters that can distinguish between different types of nevi were developed and confirmed by examining the histology of eleven of the imaged nevi. MPM features of nevi examined included cytologic morphology of melanocytes in the epidermis and dermis, the size and distribution of nevomelanocytes both within and around nests, the size of rete ridges, and the presence of immune cells in the dermis. Distinguishing features include cytological morphology, the size of nevomelanocytes, the size of nevomelanocyte nests, and the distribution of nevomelanocytes. Notably, these distinguishing characteristics were not easily appreciated in fixed tissues, highlighting essential differences in the morphology of live skin. Taken together, this work provides a morphologic compendium of normal nevi, information that will be critical in future studies directed at identifying melanocytic nevi that are evolving to melanoma.


| INTRODUC TI ON
Common melanocytic nevi, also known as "moles," are benign growths of melanocytes. Most nevi are benign, but some can be difficult to distinguish from early-stage melanoma leading to unnecessary excisions (H. K. Koh et al., 1996;Osborne, Chave, & Hutchinson, 2003). Multiphoton microscopy (MPM) is a laser scanning imaging technique that can provide label-free contrast through optical signals generated by non-linear near-infrared laser interactions with tissue molecular components. Clinical MPM is a recent and rapidly developing field of research in dermatology, holding great promise as a non-invasive imaging tool in clinical practice, particularly for early diagnosis. Our group and others have demonstrated the potential of MPM for non-invasive imaging of pigmented lesions and detection of melanoma Balu et al., 2014;Dimitrow et al., 2009;Seidenari et al., 2013). However, these studies include limited information about the MPM morphology and cytology of common nevi and no histology confirmation of the identified MPM features. Knowledge about the in vivo morphology of benign melanocytic nevi, which may differ significantly from the processed tissue morphology, is necessary for the differentiation of benign melanocytic nevi from early malignant melanomas.
In this study, we seek to explore the range of in vivo morphologies of different types of common melanocytic nevi in human skin in order to identify key MPM features of these skin lesions. Melanocytic nevi are composed of nevus cells, which differ from normal melanocytes as they are arranged in clusters or "nests" and by rarely if ever showing dendritic processes. (Lever & Schaumburg-Lever, 1990) Nevus cells lie in nests in the lower epidermis in junctional nevi and mainly in the dermis of intradermal nevi. (Lever & Schaumburg-Lever, 1990) Compound nevi possess features of both junctional and intradermal nevi.
We assess qualitatively, by comparison with histology, the presence of nests of nevomelanocytes and their cytological features in the in vivo MPM images. We also identify and confirm histologically, for the first time, the immune response in some of these melanocytic lesions. We propose several quantitative parameters related to the size of nevomelanocytes, the size of the nests, and the percentage of nevomelanocytes located outside the nests. We show that these metrics can distinguish between different types of nevi and may be relevant in assessing nevi in evolution or in distinguishing benign nevi from early-stage melanoma, or in quantifying an anti-melanocyte immune response induced by checkpoint inhibitors.

| Clinical multiphoton tomograph
We used an MPM-based clinical tomograph (MPTflex, JenLab, GmbH, Germany) for the in vivo imaging of the pigmented lesions in human skin. This imaging system consists of a femtosecond laser (Mai Tai Ti:Sapphire oscillator, sub-100 fs, 80 MHz, tunable 690-1020 nm; Spectra-Physics), an articulated arm with near-infrared optics, and beam scanning module. The imaging head includes two photomultiplier tube detectors used for parallel acquisition of twophoton excited fluorescence (TPEF) and second harmonic generation (SHG) signals. The excitation wavelength used in this study was 790 nm. The TPEF signal was detected over the spectral range of 410 to 650 nm, whereas the SHG signal was detected over a narrow spectral bandwidth (385-405 nm). We used a Zeiss objective (40×, 1.3 numerical aperture, oil immersion) for focusing the laser light into the tissue. The laser power used was 5 mW at the surface of the skin and up to 30 mW deeper into the superficial dermis of the skin. We acquired the MPM data as z-stacks of en-face images from the stratum corneum to the superficial dermis. Typically, the field of view (FOV) for each optical section was about 200 × 200 μm 2 and the step between the optical sections 5 μm. Due to the limited FOV of each individual scan, in order to sample a larger area of the lesion, we acquired several stacks of images within each lesion. Thus, a total of 2,462 images were acquired for this study, corresponding to an average of 82 images for each nevus. Images were 512 × 512 pixels and were acquired at approximately 6 s per frame. For several lesions, we acquired vertical cross-sectional, "histology-like" images from the stratum corneum to superficial dermis. These images were 1,024 × 1,024 pixels and were acquired at approximately 30 s per frame. All images were color-coded such that green and blue represent the TPEF and SHG signals, respectively. In MPM imaging of skin, the contrast mechanism is based on two-photon excited fluorescence (TPEF) signal from NADH, FAD, keratin, melanin, and elastin fibers and on second harmonic generation (SHG) signal from collagen.

| Patients
Thirty pigmented lesions in twenty-eight patients were imaged in vivo by MPM. In order to address inter-observer variability, all lesions Significance Successful translation of multiphoton microscopy to the clinical setting requires a good understanding of the live skin tissue morphology, which may differ significantly from the processed tissue morphology. This study establishes a MPM morphologic catalogue of common nevi, information that will be critical in devising strategies to distinguish them from nevi that are evolving to melanoma. An immune response in common melanocytic nevi is identified and confirmed histologically for the first time. This study proposes several quantitative parameters based on MPM features of common nevi that can distinguish between different types of nevi. that were not biopsied (19) were diagnosed as common nevi by three independent dermatologists. The pigmented lesions with a clinically suspicious appearance (11) were biopsied following MPM imaging.
The histological diagnosis was performed by three independent dermatopathologists. The topographic location and pathologic characteristics along with MPM features corresponding to each case are summarized in Table 1.
All in vivo measurements were conducted according to an approved institutional review board protocol of the University of California, Irvine (HS No. 2011-8494), with written informed consent obtained from all patients.

| Image analysis
We used ImageJ for both the qualitative and quantitative analyses.
The qualitative analysis included: (i) basic processing tasks of the MPM images for enhancing contrast and highlighting features of interest; and (ii) comparison of the MPM and histologic images of biopsied nevi in order to identify and correlate their morphological features.
The quantitative analysis included measurements of: (1) nevomelanocyte sizes-We calculated the average cell diameter using a semi-automatic segmentation method within ImageJ. We used the following procedure for nevomelanocyte segmentation: Gaussian

| RE SULTS
We analyzed both qualitatively and quantitatively the MPM images acquired in vivo in thirty common nevi in human skin (see Table 1 for a summary of the lesion location and diagnosis). Besides the enhanced pigmentation of keratinocytes, which may also be present and imaged in pigmented normal skin (Balu et al., 2014;Saager et al., 2015),

| Nests of nevomelanocytes
The presence or absence of nests was consistent in all the stacks acquired in 60% of the patients. In the rest of the patients, the nests of nevus cells were imaged on average in 49% of the total num-

| Size of the nevomelanocyte nests
There was no significant difference between the size of nevomelanocyte nests in the two groups of junctional/compound and intradermal nevi. In all nevi, the size of the vast majority of the nests (~90%) was between 20 and 80 μm and did not significantly change with the imaging depth. Nine of the thirteen largest nests were located in intradermal nevi and had sizes within the range of 85-170 μm. This analysis included 78 nests visualized in 13 junctional/compound nevi and 62 nests in 6 dermal nevi. Figure 3 shows the distribution of the nevomelanocyte nest diameters measured in the two groups of nevi (junctional/compound versus dermal) for different patients.

| Size of nevomelanocytes
The nevomelanocytes were significantly smaller (p = .0001) in junctional and compound nevi (8.94 ± 2.9 μm) compared to the nevomelanocytes in the intradermal nevi (14.08 ± 3.3 μm). Figure 4 shows the distribution of the nevomelanocyte diameters for the two groups of nevi for different patients.

| Percentage of nevomelanocytes outside the nests
Most of the nests (76%) were rather small including less than 45 nevomelanocytes. The nests with the largest number of nevomelanocytes (100-125) were located in intradermal nevi. In most nevi (72%), the nevomelanocytes were located exclusively inside the nests.
The rest of nevi showed nevomelanocytes located both within and F I G U R E 3 Distribution of nevomelanocyte nest diameters in junctional/compound and intradermal nevi. Data represent individual nest diameter measurements for each lesion. Case numbers correspond to those listed in Table 1 outside nests. The number of sparse melanocytes outside nests represented less than 5% of those located within the nests.
Elongated rete ridges represented another specific feature we imaged in nevi. We identified this feature exclusively in junctional nevi.
We compared the elongation of the rete ridges based on their morphology in the en-face images. At the dermo-epidermal junction, the enface MPM images show a horizontal sectional view of the rete ridges.
In this view, the rete ridges appear as ring-like structures of basal keratinocytes surrounding the dermal papilla. In the nevi with long rete ridges, this morphology was visualized as deep as ~ 150 μm, the common depth limit for MPM imaging of nevi. For the nevi with shorter rete ridges, the en-face images captured at ~ 150 μm depth showed the presence of collagen and elastin fibers, that is, the morphology characteristic to the skin dermis. Representative images that illustrate the assessment of rete ridges are included in the Supporting Information, in Figure S4. We identified elongated rete ridges in 6 out of 11 junctional nevi. Nests of nevomelanocytes were not visualized in 5 of these nevi, likely due to their deeper location at the tips of the rete ridges.

| Immune response
We captured the immune response in the form of dermal infiltration of inflammatory cells and melanophages. The presence or absence of the immune response was consistent in all the stacks acquired in 80% of the patients. In the rest of the patients, the inflammation was captured on average in 53% of the total number of stacks acquired.
Multiphoton microscopy imaging revealed the presence of inflammatory cells in two of the nevi imaged in vivo by MPM. These nevi were histologically diagnosed, one as residual dermal melanocytic nevus (Case 20), the other one as recurrent compound melanocytic nevus (Case 27). Figure 5 (b-e) shows dermal cells visualized by MPM in vivo in the recurrent compound melanocytic nevus (Case 27), along with a nest of nevomelanocytes at the dermo-epidermal junction. The appearance of dermal cells is different from the nevomelanocytes imaged in this lesion or in others (Figures 1, 2, S2, Video S1, Video S2), and based on correlation with immunohistochemistry (Figure 5b), they are most likely CD3 T cells. In the MPM images, these cells were approximately round in shape, were about 10-12 μm in size, and arranged in clusters. Unlike nevomelanocytes, they appear dispersed rather than closely packed and their nuclei are not well resolved.
Multiphoton microscopy imaging also revealed the presence of melanophages in the dermis of eight nevi. Six of these nevi were biopsied and histologically diagnosed as compound (2) and intradermal (4) nevi. In the MPM images, melanophages had an irregular shape, were about 18-20 μm in size, and were generally bright due to fluorescence from their abundant melanin content. They were arranged F I G U R E 4 Distribution of nevomelanocyte cell diameters in junctional/compound and dermal nevi. Data represent number of cells with different cell diameters measured in junctional/compound nevi (a) and intradermal nevi (b) for each lesion. Case numbers correspond to those listed in Table 1 in clusters, and some of them appeared dendritic. Their appearance correlated well with the melanophages identified in the histological images. Representative example of MPM images showing melanophages imaged in an intradermal nevus is shown in Figure 6 (c,d, Case 23). The presence of melanophages in this lesion was confirmed by histology (Figure 6b).
Establishing the cytological morphology of nevomelanocytes and immune response by comparison with histopathology allowed us to accurately identify these cells in nevi that were not biopsied.
Representative MPM images of nevomelanocytes and melanophages imaged in such nevi are shown in Figures S2 and S3, respectively (Supporting Information).

| D ISCUSS I ON
In this study, we assessed both qualitatively and quantitatively the cytological and morphological features of common nevi in human skin as visualized by in vivo MPM.
Nests of nevomelanocytes represented the most predominant feature captured by MPM in most nevi. We assessed several quantitative parameters related to the size of nevomelanocytes, the size of the nests, and the percentage of nevomelanocytes located outside the nests. The nevomelanocytes of intradermal nevi were significantly larger and presented larger nuclei than the nevus cells in junctional and compound nevi. Their appearance correlated well with the corresponding histological images and, based on their large size and epithelioid shape, seemed to represent the superficial component (type A melanocytes) of so-called "maturation" (Massi & LeBoit, 2013). This term is commonly used in histology to indicate the progression of nevomelanocytes from large epithelioid cells in the papillary dermis to smaller, lymphocyte-like cells in the superficial dermis and spindle Schwann-like cells deeper in the dermis. In all nevi, the size of the vast majority of the nests (~90%) was in the range of 20-80 μm and did not significantly change with the imaging depth.
The nevomelanocyte nest size can be a potential relevant metric in differential diagnosis of benign and malignant melanocytic nevi. (Pellacani, Cesinaro, & Seidenari, 2005) The limited imaging depth of MPM in skin (~200 μm) did not allow us to measure a potential change in size of the nevomelanocyte nests with depth, but we were able to show that the majority of melanocytes were located within the nests. Published studies have suggested that melanocytes within nests express the senescence marker p16 (Koh & Cassarino, 2018).
This observation indicates that the metric related to percentage of nevomelanocytes outside the nests may be used to assess the nevi in evolution. Metrics based on size of nests and/or the percentage of cells located outside nests may also be used as criteria to distinguish nevi from melanoma. This hypothesis is based on observations from histology (Price, Rywlin, & Ackerman, 1976) and from our previously published work on MPM imaging of nevi and melanoma (Balu et al., 2014), but it needs to be tested in further studies.
Elongated rete ridges, a known histological feature of junctional nevi (Massi & LeBoit, 2013), were imaged by MPM in about half of these types of nevi. They generally hindered the imaging of nevomelanocytes nests. Thus, while nests of nevomelanocytes were imaged in most of the nevi, they were not captured in most nevi that showed elongated rete ridges. This was likely due to the location of the nevomelanocytes at the tips of the rete ridges (Massi & LeBoit, 2013) that were too deep to access by MPM. Limited penetration depth of MPM was also likely the reason we did not capture the nevomelanocytes in three out of eight nevi diagnosed as predominantly intradermal.
Inflammatory infiltrates may develop in some types of melanocytic lesions such as recurrent and residual melanocytic nevi that involve regression (Fox, Reed, & Shea, 2011). MPM was able to capture dermal cells with a distinct morphology in a regressing nevus in approximately the same location where infiltrating CD3 + cells were observed on histology. T-cell infiltration in nevi is often a sign of regression or of an anti-melanocyte immune response and can be indicative of a good response to immunotherapy (Linck, Costa, & Garicochea, 2017 (Balu, Mikami, Hou, Potma, & Tromberg, 2016).
In conclusion, this study illustrates and discusses qualitative and Tromberg are co-authors of a patent owned by the University of California, which is related to the technology described in this study.
The Institutional Review Board and Conflict of Interest Office of the University of California, Irvine, have reviewed patent disclosures and did not find any concerns. No potential conflicts of interest were disclosed by the other authors.