Reflectance confocal microscopy terminology glossary for melanocytic skin lesions: A systematic review

Abstract

Background:

There is lack of uniformity in the reflectance confocal microscopy (RCM) terminology for melanocytic lesions.

Objective:

To review published RCM terms for melanocytic lesions and identify redundant, synonymous terms.

Methods:

Systematic review of original research articles adhering to PRISMA guidelines was conducted until August 15, 2018. Two investigators gathered all published RCM terms used to describe melanoma and melanocytic nevi. Synonymous terms were grouped based on similarity in definition and in histopathological correlation.

Results:

Out of 156 full-text screened articles, 59 studies met the inclusion criteria. We identified 209 terms; 191 (91.4%) corresponding to ‘high-magnification/cellular level’ terms and 18 (8.6%) corresponding to ‘low-magnification/architectural patterns’ terms. The overall average use frequency of RCM terms was 3.1 times (range 1 – 31). By grouping of individual RCM terms based on ‘likely-synonymous’ definitions and by eliminating terms lacking clear definition, the total number of RCM terms could be potentially reduced from 209 to 40 terms (80.8% reduction).

Limitations:

Non-English and non-peer reviewed articles were excluded.

Conclusions:

This systematic review of published RCM terms identified significant terminology redundancy. It provides the basis for subsequent terminology consensus on melanocytic neoplasms.

INTRODUCTION

Reflectance confocal microscopy (RCM) allows for non-invasive in vivo visualization of the skin at a “quasi-histological” resolution.¹ In the 1990s, the histological correlates of RCM attributes of skin were first described.^{2, 3} Since, numerous publications have shown the utility of RCM in the diagnosis of melanoma^4–7 and non-melanoma skin cancer,^8–10 in monitoring of non-invasive skin treatments^{11, 12} and in guiding dermatological surgery.^13–16

As demonstrated in a previous systematic review, RCM terminology in the literature lacks consistency and contains many redundant terms.¹⁷ We identified a total of 139 RCM terms used to describe non-melanocytic lesions (NMLs); by grouping these terms for synonymy, we were able to further shorten the list of terms by more than 50%.¹⁷ For melanocytic lesions, a consensus study of RCM terminology was published in 2007 by Scope et al;¹⁸ however, a multitude of new RCM terms, describing melanocytic neoplasms, has been published since.

Recently, category I Current Procedural Terminology (CPT) reimbursement codes have been allotted to RCM imaging in the U.S;^{1, 19} we anticipate a consequent rise in the integration of RCM into clinical practice.²⁰ To that end, there is a pressing need for standardization of RCM terminology of melanocytic lesions^{17, 21} – to enable structured reporting of RCM examinations and to facilitate RCM teaching to novices. Herein, we performed a systematic review of the terms used to describe the RCM features of melanocytic neoplasms and identified redundant, synonymous terms.

METHODS

The results of this systematic review were obtained according to the guidelines for reporting systematic reviews as published in the PRISMA Statement (available in www.prisma-statement.org). All images used for illustrating RCM terms were acquired using a commercial RCM system (Vivascope 1500 or Vivascope 3000, Caliber ID, Rochester, NY), under an IRB protocol (#17–083). The principles of RCM imaging have been previously described.¹

Eligibility criteria

We included all original, peer-reviewed RCM articles published between 1995 and 2018 that contained the diagnosis of nevi, including congenital and Spitz nevi, and cutaneous melanoma. So-called “borderline” melanocytic lesions, such as atypical Spitzoid tumors, were also included. We excluded articles describing RCM features of collision tumors and melanocytic lesions of special anatomic sites, such as genitalia, nails and eyelids. Literature reviews, single case-reports, conference abstracts, animal studies, and publications lacking full-text were excluded; due to the lack of peer-review process, we also excluded book chapters.

Information sources, search and Study selection

Systematic literature searches were conducted (August 15, 2018) in four databases with no specified date, age, sex, or language restrictions. The databases searched were: (1) MEDLINE (via PubMed); (2) Embase; (3) The Cochrane Library (Cochrane); and (4) Web of Science (WoS). In an effort to be comprehensive and include grey literature publications into the data set of citations, conference proceedings and abstracts were retrieved from Embase and WoS utilizing broad and inclusive publication-type filters. Search results were combined in a bibliographic management tool (EndNote, Clarivate Analytics) and duplicates were eliminated both electronically and manually to ensure an efficient de-duplication process. The search strategy employed the Medical Subject Headings (MeSH) phrases: [“Microscopy, Confocal”] AND [“Skin Neoplasms” OR “Dermatology” OR “melanoma” OR “nevus” AND “in vivo”] AND [“Terminology” OR “Current Procedural Terminology” OR “Terminology as Topic” OR “Dictionaries as Topic” OR “Data Accuracy” OR “Algorithms” OR “Reproducibility of Results” OR “Classification”]. Bibliographies within retrieved articles were also reviewed to identify additional studies. For this specific systematic review, we included only RCM terms pertinent for melanocytic lesions and excluded terms related to non-melanocytic neoplasms (e.g., basal cell carcinoma, squamous cell carcinoma and lichen planus-like keratosis). Two authors (C.N-D. and K.L.) independently screened all relevant titles and abstracts for eligibility. If necessary, full-text articles were screened. Differences in judgment were resolved with a third reviewer (M.J.) until consensus was achieved (Figure 1).

Figure 1:

PRISMA diagram.

Data collection and extraction process

Two authors (C.N-D. and K.L.) extracted data from the included studies independently. Disagreements were resolved by consensus; if agreement could not be reached, a third author (M.J.) was used as referee. The following information was extracted from each study: the RCM term, its definition, diagnosis associated with the term, the histopathological correlates associated with that term (when reported), and whether the term was used in an algorithm. All extracted RCM terms were recorded as published in the literature, chronologically, in an Excel spreadsheet (Microsoft, Redmond, WA). To weight the use-frequency of RCM terms, we recorded the number of studies that utilized each term. Lastly, five authors (C.N-D., K.L., J.M., S.A., M.J.) grouped all terms into ‘likely-synonymous’ terms based on their RCM definitions and on similar histopathologic correlates (e.g. “pagetosis’, “pagetoid infiltration”, and “atypical cells in the epidermis”). Next, the ‘likely-synonymous’ RCM terms were organized by the skin layer to which they had been ascribed: (1) epidermis, (2) dermo-epidermal junction (DEJ), and (3) dermis. Finally, we categorized all ‘likely-synonymous’ RCM terms into two categories: (a) high-magnification, cellular-level terms (based on optical sections, typically 0.5×0.5μm to 1×1mm in area) (e.g. atypical cells, inflammatory infiltrate), and (b) low-magnification patterns (based on mosaic RCM images, typically >1×1mm to 8×8mm in area) (e.g. ringed, meshwork or clod pattern). RCM terms lacking definition in the main text, tables, or figures of an article (or where the definition could not be clearly inferred by the name of the term), were listed as ‘definition not available (N/A)’. Additionally, following previously-published approach for simplifying terminology,²² we extracted basic terminology units from composite term (e.g. the term “Large roundish cells in the epidermis” contains the basic units “cells”, “large”, roundish” and “epidermis”). ‘Composite’ terms were defined as those that contained two or more basic units, while ‘simple’ terms were those with a single unit. We then categorized the basic elements into: ‘cells / structure’ (the cell type could be explicit or implied by the term, e.g., “cells”–in the aforementioned example implies melanocytes); ‘morphological descriptors / modifiers’ (e.g., “large” and “roundish”); and ‘anatomic location / distribution (e.g. “epidermis”).

Summary measures and statistical analysis

Descriptive statistics were used to categorize the number of RCM terms by diagnosis and by anatomic layer. ‘Use frequency’ was based on the number of papers describing each RCM term. ‘Average use frequency’ described the proportion between ‘use frequency’ for RCM terms (as individual terms or as subgroups of likely-synonymous terms) divided by the total ‘use frequency’ for all terms describing that diagnosis.

RESULTS

After screening 156 full-text articles, 59 studies were identified that met the inclusion criteria (Figure 1).^{4–6, 23–78} In total, 209 RCM terms were identified: 191 (91.4%) ‘high-magnification/cellular-level’ terms, and 18 (8.6%) ‘low-magnification/architectural pattern’ terms. The use frequency of each individual RCM term and ‘likely-synonymous’ groups of RCM terms are shown in Tables I–II. Representative images for each category are shown (Figures 2&3).

Table I:

Reflectance confocal microscopy terms used to describe melanocytic lesions at high magnification, cellular-level resolution

RCM term	Use frequency of RCM term, n	Definition	Part of an algorithm	Histopathologic correlates
EPIDERMIS (n=68; 35.4%)
- Terms describing presence of cells in the epidermis, utilizing ‘pagetoid’ without a specified cell shape
Pagetoid cells4, 6, 23, 28–30, 35, 39, 41, 42, 48, 49, 52, 55, 58, 61, 63–68, 72, 75–78	27	Large nucleated cells, twice the size of keratinocytes (>20 um), with a dark nucleus and bright cytoplasm.		Presence of melanocytes in suprabasal layers of the epidermis
Pagetoid spread25, 32, 33, 56–58, 60, 62, 75	9		Borsari 201825
Widespread pagetoid infiltration4, 6, 36, 37, 43, 53, 58, 59,68	9		Pellacani 200768
Pagetoid melanocytes44, 54	2
Cells in pagetoid pattern70	1
Pagetoid spread of atypical melanocytes27	1
Irregular intraepidermal growth of melanocytes at the periphery of the lesion32	1
-Terms describing presence of cells in the epidermis, specifying ‘pleomorphic (round and dendritic) shape
Polymorphic melanocytic cells45–47	3	Variability of the aspect of pagetoid cells (round and dendritic)
Pleomorphic pagetoid infiltration68	1
Pleomorphic pagetoid cells23	1
Pleomorphic pagetoid shape56	1
Pleomorphism (in the epidermis)72	1
Bright dendrites/Dendritic/Round cells within the epidermis64	1
Striking pleomorphism49	1
-Terms describing presence of cells in the epidermis, specifying ‘round’ shape
Roundish pagetoid cells36, 37, 49, 56, 59, 68, 72, 74, 75	9	Large roundish nucleated cells, with a dark nucleus and bright cytoplasm, within suprabasal layers	Pellacani 2007,68 Segura 200923
Round pagetoid cells4, 6, 52, 53, 57, 58	6		Guitera 20104
Round cells5, 38, 57	3
Large roundish cells in the epidermis34	1
Round pagetoid cells with halo4*	1
-Terms describing presence of cells in the epidermis, specifying ‘dendritic’ shape
Dendritic pagetoid cells4, 6, 49, 50, 52, 56, 58, 68, 72, 74, 75	11	Bright nucleated cells with dendritic-like branches within suprabasal layers
Dendritic cells5, 38, 55, 57	4	--
Large dendritic cells in the epidermis34	1	--
-Terms describing presence of cells in the epidermis, specifying ‘low refractivity’
Hyporeflective pagetoid cells77, 78	2	Round dark structures (similar to ‘holes’) within the epidermis
Dark pagetoid cells52*	1	--
-Terms describing presence of ‘atypical’/’irregular’ cells in the epidermis without a specific shape
Cell atypia6, 52, 58	3	--
Atypical cells24, 69	2	--
Enlarged atypical melanocytes32	1	--
Irregularly bright melanocytic cells46	1	--
Weighted subtotal	105
-Terms describing the presence of dendrites in the epidermis
Composite branching dendrites37, 46, 47, 55	4	Numerous bright tangled lines within the epidermal layers, originating from dendritic cells with not always visible cell body		Dendritic projections of melanocytes or Langerhans cells in the epidermis
Simple branching dendrites46, 47	2
Bright dendrites/Tangled lines/dendritic cells64	1
Dendritic structures26	1
Dendritic processes26	1
Frequent, coarse, branching dendrites54	1
Weighted subtotal	10
-Terms describing the presence of melanocyte infiltration of hair follicles
Follicular localization of pagetoid cells and/or atypical cells4, 29, 34, 61	4	Infiltration of dendritic/roundish cells in the inner portion of the hair follicle or elongated protrusions around the entire perimeter of the follicle	Guitera 20104	Infiltration of hair follicles and adnexal structures by atypical melanocytes. Typically seen in melanoma of the lentigo maligna type
Pagetoid cells around follicular opening6, 52, 58	3
Folliculotropism24, 64	2
Atypical melanocytes surrounding	1
adnexal openings26
Infiltration of adnexal structures27	1
Irregular aggregates of bright perifollicular cells31	1
Atypical cells infiltrating follicular structures62	1
Weighted subtotal	13
-Terms describing the presence of a normal/regular spinous and granular layer:
Honeycombed pattern (typical/regular)4, 23, 38, 39, 43, 49, 55–57, 63, 65, 66, 68, 75	14	Well-demarcated cellular outlines of keratinocytes forming a grid-like structure resembling a ‘honeycomb’ in the spinous and granular layers.		Normal epidermal keratinocytes
Readily detected keratinocyte cell borders45–47	3
Regular epidermal pattern67	1
-Terms describing the presence of a normal/regular basal layer:
Cobblestone pattern4, 23, 38, 39, 42, 43, 49, 57, 58, 63, 65, 66, 68, 69, 72, 74, 75, 77	18	In the basal layer aggregates of small, nonnucleated polygonal cells and bright cytoplasm can be seen creating a ‘cobblestone’ pattern.
Monomorphic melanocytic cells45–47	3
Cobblestone with small nucleated cells72, 74	2
Typical cells in the basal layer28, 36	2
Homogeneous bright melanocytic cells46	1
Monomorphic reflactile cells55	1
Typical basal cells23	1		Segura 200923
Weighted subtotal	46
-Terms related to acanthosis of the epidermis
Broadened honeycomb pattern4, 23, 58, 61, 69, 72, 74, 75	7	Prominent bright epidermis intermingled with papillae		Acanthosis of the epidermis
Acanthosis49	1
Weighted subtotal	8
-Terms describing with the presence of atypical keratinocytes/ disarrangement of keratinocytes
Epidermal disarray4, 6, 23, 27, 50, 52, 55, 58, 68	9	Irregularly shaped keratinocytes with poorly defined or absent borders		Atypical keratinocytes with variation in size, shape, and crowding of the nuclei
Atypical honeycomb58, 68, 69, 72, 74, 77	6
Disarranged epidermis57, 67, 69, 72, 75	5
Disarray38, 56, 65, 66	4
Disarranged epidermal pattern43, 63, 67	3
Poorly defined or absent keratinocyte cell borders45–47	3
Irregular epidermal pattern67	1
Honeycomb atypical and disarray52	1
Irregularly shaped keratinocytes43	1
Poorly defined keratinocyte cell border54	1
Loss of keratinocyte cellular borders54	1
Honeycombed and irregular keratinocytes64	1
Epidermal disruption49	1
Weighted subtotal	37
-Terms related to a ‘speckled’ appearance in the epidermis
Grainy image54, 55, 68, 72	4	Bright granular dust-like particles barely discernible as individual granules at the level of epidermis		Extracellular melanin granules in the epidermis
Epidermal granularity57	1
Weighted subtotal	5
DERMOEPIDERMAL JUNCTION (n=46; 23.9%)
-Terms associated with an irregular papillary contours
Non-edged papillae4, 6, 23, 29, 31, 33, 36, 37, 39–41, 43, 48, 49, 52, 53, 57–59, 61, 63, 65, 67–70, 72, 74–77	31	Dermal papillae without a demarcating bright rim of cells; the inter-papillary areas of the epidermis are widened and often show large atypical melanocytes	Pellacani 200768 and Guitera 20104	Disarranged rete ridges with a disorderly proliferation of melanocytes not confined to the sides and tips of the rete ridges. Loss of the papillary contour is associated with flattening of the DEJ.
Irregular ringed30	1
-Terms associated with loss of papillae contours
Non-visible dermal papillae4, 50, 58, 75	4
Irregular / disarrayed dermal epidermal junction39, 44, 64	3
Non-visible papillary contours40, 69	2
Loss of dermal papillae (flat DEJ)31	1
Poorly visualized dermal papillae54	1
DEJ disarray57	1
Disarrangement at the DEJ77	1
Absence of papillae77	1
Weighted subtotal	46
-Terms associated with regular dermal papilla
Edged papillae4, 6, 23, 28, 36, 48–50, 52, 57, 58, 63, 65, 67–72, 74, 77	22	Dermal papillae with clearly outlined contours. They may result from single cells forming rims around the papillae (‘rings’), or by ‘junctional nests’ constituted by compact melanocytic aggregates with sharp borders	Segura 200923	Pigmented basal keratinocytes and melanocytes along the sides of the rete ridges
Rings64	1
Weighted subtotal	23
-Terms describing the presence of ‘atypical’ cells at the DEJ
Atypical cells at the DEJ29, 33, 39, 42, 48, 49, 67, 70, 76, 78	10	Presence of bright round or dendritic nucleated cells that are abnormally large in size (>50 um), display unusual contour (e.g. Triangular, starshaped) or have large and eccentric nuclei at the DEJ		Proliferation of atypical melanocytes as solitary units along the DEJ
Cytological atypia25, 30, 41, 53, 56, 59, 60, 63, 65	9		Borsari 201825
Marked atypia of basal cells6, 23, 52, 58, 72	5
Atypical cells40, 68	2
More than 3 atypical cells at the junction in five images4	1		Guitera 20104
Marked cell atypia in the basal layer75	1
Bright nucleated cells26	1
Scattered bright structures31	1
Predominance of single cells over nests32	1
Large dendritic or round cells at the epithelial-stromal junction and/or in the stroma35	1
Atypical nucleated cells at the DEJ36	1
Cellular atypia in the DEJ37	1
Atypical cells at basal layer43	1
Mild and marked cellular atypia4	1
Atypical and pleomorphic refractile cells55	1
Bright, highly refractile particles55	1
Presence of large visible cells68	1
-Terms describing atypical cells with other shapes at the DEJ
Spindled cells49, 57, 69**	3
-Terms describing presence of ‘atypical’ cells with additional features
Focal increase of atypical melanocytes and nests27	1
Weighted subtotal	43
-Terms describing the presence of round-shaped junctional aggregates, regular
Junctional nests5, 40, 48, 49, 55–58, 67–69, 72, 74, 77, 78	15	Elongated and branching tubular structures with heterogeneous reflectance harboring aggregates and individual polymorphous nucleated cells at the DEJ. They often bulge into dermal papillae.		Junctional nested proliferation of melanocytes. Elongated junctional nests of melanocytes may bridge between adjacent rete ridges
Junctional clusters4, 23, 58, 63, 65, 68, 71, 72	8
-Terms describing the presence of elongated, tubular junctional aggregates
Junctional thickening4, 26, 43, 58, 63, 65, 68, 69, 71, 72, 74	11
Cord-like rete ridges27	1
Nonhomogeneous cellularity67	1
- Terms describing the presence of short/bridging tubular junctional aggregates
Mitochondria-like structures77	1
Short interconnections67	1
Weighted subtotal	38
-Terms describing sheet-like proliferation of melanocytes at the DEJ
Sheet-like structures5, 23, 24, 43, 49, 56, 67–69, 72, 74, 77	12	Cells distributed at the transition of the DEJ showing loss of dermal papillae not aggregated in clusters but closely distributed in the same plane with the loss of dermal papillae		Florid lentiginous proliferation of atypical melanocytes along the DEJ; mostly seen in melanomas in sun-damaged skin
Sheet of cells4, 6, 30, 52, 58	5
Sheet-like25	1
Sheets of round to dendritic nucleated cells26	1
Tangled filaments/dendrites crossing the papillae41	1
Sheets of mainly dendritic atypical cells27	1
Weighted subtotal	21
-Terms describing elongated structures bulging from the hair follicles
Medusa head-like24, 64	2	Small protrusions or elongated protrusions from hair follicles. They correspond to the “medusa head-like” structure when distributed around the entire perimeter of the follicle		N/A
Bulging around hair follicle64	1
Weighted subtotal	3
SUPERFICIAL DERMIS (n=68; 32.8%)
-Terms describing the presence of solitary melanocytes in the papillary dermis
Nucleated cells within the dermal papillae4, 6, 28, 29, 36, 37, 52, 59, 61, 68, 76	11	Round to oval or triangular cells with welldemarcated bright cytoplasm and well demarcated dark nucleus infiltrating dermal papillae	Guitera 20104	Presence of atypical melanocytes in the superficial (papillary) dermis
Nucleated cells in the dermis58, 69, 72, 75	4
Isolated cells in the papilla72, 74	2
Dermal infiltration of roundish cells64	1
Atypical melanocytes within the upper dermis27	1
Atypical cells in the dermal papilla33	1
Dermal nucleated cells57	1
Isolated cells within the upper dermis63	1
Isolated cells within the papillary dermis65	1
Atypical nucleated cells in dermis67	1
Atypical nucleated dermal cells23	1		Segura 200923
Atypical cells infiltrating papillary dermis78	1
Spindle cells in superficial dermis58***	1
Weighted subtotal	27
-Terms describing the presence of cohesive nests of melanocytes in the papillary dermis
Dense nests4, 25, 48, 49, 56, 58, 60, 67–72, 77	14	Compact aggregates with sharp margins and monomorphous cells in which it is difficult to discern individual cell borders	Borsari 201825	Round to oval junctional or dermal nests of melanocytes
Dermal nest5, 23, 40, 43, 48, 55, 57, 67, 78	9
Melanocytic cell nests45–47	3
Dense clusters63, 65, 74	3
Dense regular nests23, 75	2
Confluence of nests69**	1
Dermal cell clusters72	1
Nesting25	1
Weighted subtotal	34
-Terms describing the presence of discohesive/irregular nests of melanocytes in the papillary dermis
Dense and sparse nests25, 30, 48, 49, 56, 60, 67, 70	8	Clusters of somewhat loosely aggregated cells in which some of cells have clearly defined cell borders and little dark space can be seen inbetween some, but not all, cells		Irregular or discohesive nests of melanocytes
Dishomogeneous nest4, 6, 52, 58, 69, 72, 74	7
Sparse nest58, 68, 69, 72, 74	5
Non-homogeneous nests23, 68	2
Irregular clods30	1
Atypical nucleated cells arranged in nests41	1
Aggregates of atypical cells62	1
Dishomogeneous clusters63	1
Sparse cell clusters65	1
Marked pleomorphism within nests49	1
Discohesive junctional nests77	1
Dense irregular nests75	1
Dense dishomogeneous clusters75	1
Weighted subtotal	31
-Terms describing the presence of ‘cerebriform’ aggregates in the dermis
Cerebriform nest5, 6, 23, 30, 36, 37, 40, 48, 52, 53, 56, 58, 68, 69, 72, 74–76	18			Nodular aggregates of atypical melanocytes in melanoma with dermal component
Cerebriform clusters49, 59, 63, 65, 75	5
Weighted subtotal	23
-Terms describing the presence of melanophages in the papillary dermis
Plump cells23, 33, 56–58, 63, 69, 70, 72–75, 77	13	Irregularly shaped bright cells with ill-defined borders and usually no visible nucleus		Melanophages
Melanophages25, 26, 30, 50, 51, 60, 64, 65	8		Borsari 201825
Plump-bright cells4, 40, 68	3
Bright, round-to-triangular, nonnucleated cells26	1
Weighted subtotal	25
-Terms describing the presence of inflammation (other than melanophages) in the papillary dermis
Inflammation38, 49	2	Bright spots and small bright particles in the dermis		Inflammatory cells (other than melanophages)
Bright spots56, 69	2
Inflammatory infiltrate48, 67	2
Dermal inflammation24	1
Dermal bright cells51	1
Bright small cells and/or hyperreflective spots68	1
Small bright cells and particles72	1
Bright particles73	1
Bright hyperreflecting spots23	1
Bright dots77	1
Weighted subtotal	13
-Terms describing patterns of collagen in the papillary dermis
Bundled collagen72, 74, 77	3	Elongated fibrillar structures (1 – 5 um) without cellular component distributed side by side through the dermis		Various correlates ranging from normal collagen scar-like collagen, to solar elastosis.
Reticulated collagen72, 74, 77	3
Coarse collagen24, 70	2
Bright fibrillar structures75	1
Thickened collagen / curled fibers (elastosis)64	1
Stromal fibre (‘collagen’) morphology48	1
Highly reflecting fibers50	1
Curled highly refractive collagen fibers4	1
Fibroplasia67	1
Reticulated fibers68	1
Broadened reticulated fibers68	1
Thick cordons68	1
Weighted subtotal	17
-Terms describing the presence of prominent blood vessels in the papillary dermis
Enlarged vessels23, 50, 75	3	Blood vessels appear as dark tubular structures in the dermis in which movement of bright round cells (white blood cells) is seen. Irregular vessels refer to blood vessels with abnormal diameter, density, or orientation compared to normal skin		Dilated or increased vascularity in the superficial dermis
Tortuous morphology of vessels33	1
Irregular vessels40	1
Prominent vascularity56	1
Horizontal vessels58	1
Atypical vessels59	1
Weighted subtotal	8
TERMS WITH DEFINITION NOT AVAILABLE (N=9; 4.7%)
Nests (NOS)4, 24, 58, 64, 69	5	--		--
Disarray of melanocytic architecture45–47	3	--		--
Ill-defined follicle contour64	1	Aspect of the external border of the hair follicle, defined according to its outline		--
Atypical cells with dark halo4*	1	--		--
Oval cells57	1	--		--
Elongated cells57	1	--		--
Triangular cells57	1	--		--
Cord-like structures62	1	Pseudonetwork on dermoscopy		--
DEJ atypia57	1
N/A TOTAL	15
TOTAL TERMS (N=191; 100%)

^*Described in amelanotic/hypomelanocitc melanoma.

^**Described in Spitzoid lesions.

^***Described in desmoplastic melanoma.

Table 2:

Reflectance confocal microscopy terms used to describe the patterns of melanocytic lesions at low magnification.

RCM term	Use frequency of RCM term, n	Definition	Histopathologic correlates
TERMS WITH DEFINITION (N=15)
Ringed pattern5, 24, 25, 38–42, 48, 49, 56, 62, 67, 70, 71	15	A pattern consisting of edged-papillae presenting a demarcated rim of bright basal cells forming ‘rings’	Lentiginous or small-nested junctional proliferation of melanocytes
Weighted subtotal	15
Meshwork pattern5, 24, 25, 30, 38–42, 48, 56, 62, 67, 70, 71, 77	16	Enlarged interpapillary spaces predominantly constituted by junctional thickenings and/or non-edged papillae	Nested junctional proliferation of melanocytes with bridging between adjacent rete ridges
Weighted subtotal	16
Clod pattern25, 39–42, 48, 56, 67, 70, 71	10	A low-magnification pattern composed of predominance of dense compact nests/clusters of melanocytes within the superficial dermis	Dermal nested proliferation of melanocytes
Small clods (<150 μm)70	1
Large clods (>150 μm)70	1
Weighted subtotal	12
Mixed pattern70	1	Combination of any of the DEJ patterns (i.e. ringed, meshwork or clod pattern).	Melanocytic neoplasms with a junctional and dermal components
Weighted subtotal	1
Non-specific25, 30, 39, 41, 48, 56, 67	7	Lack of a recognizable pattern at low-magnification mosaic view of the DEJ (i.e. absence of ringed, meshwork or clod pattern). Usually associated with abrupt or vague epidermadermal transition	Melanocytic proliferation with a flattened DEJ or marked attenuation of the undulation DEJ pattern
Structureless area38	1
Aspecific pattern60	1
Weighted subtotal	9
Asymmetry47	1	The overall distribution of confocal structures in one half of the lesion does not mirror those in the other half	Asymmetry
Architectural disorder41	1
Weighted subtotal	2
Peripheral rim of nests48	1	Presence of clusters of cells (nests) detectable along the entire perimeter of the lesion	Presence of junctional or dermal nests of melanocytes at the periphery of the lesion
Peripheral melanocytic nests62	1
Rim of nests at the periphery69	1
Weighted subtotal	3
Sharp border cut-off69	1	Clear demarcation between lesional and peripheral skin. Seen in Spitz nevus.	Sharp demarcation
Weighted subtotal	1
TERMS WITH DEFINITION NOT AVAILABLE (N=3; 10.3%)
Uneven pattern24	1		--
Undefined epidermal pattern43	1		--
Specific pattern60	1		--
Weighted subtotal	3
TOTAL FOR PATTERNS TERMS (N=18)	62

Figure 2:

Examples of reflectance confocal microscopy terms. A-D. ‘High-magnification’/cellular details examples of melanoma features. A. Round (red arrows) and dendritic (yellow arrows) pagetoid cells in the epidermis (2.0 × 2.0 mm). B. Atypical cells at the dermoepidermal junction (DEJ) (yellow arrows) and non-edged papillae (blue asterisks) (2.0 × 2.0 mm). C. Bright, large, dendritic nucleated cells (yellow arrows) with folliculotropism (yellow asterisk shows a hair follicle) in a lentigo maligna melanoma (0.75 × 0.75 mm). D. Dishomogeneous (‘dense and sparse’) nest (yellow arrows) at the dermis level (1.0 × 1.0 mm). E-H. ‘Low-magnification’ / patterns examples used to describe melanocytic lesions. E. Ringed pattern (2.0 × 1.0 mm). F. Meshwork pattern (2.7 × 2.0 mm). G. Clod pattern (2.5 × 3.5 mm). H. Non-specific pattern.

The overall average use-frequency of the RCM terms was 3.1 times (653 mentions of 210 terms; range 1–31 mentions). For ‘high-magnification’ terms, average use-frequency was 3.0 times (591 mentions of 191 terms; range 1–31), while for ‘low-magnification’ terms, average use-frequency was 3.4 times (62 mentions of 18 terms; range 1–18). The most-commonly used high magnification terms were ‘non-edged papillae’ (use frequency, n=31), ‘pagetoid cells’ (n=27), and ‘edged papillae’ (n=22). Only 14 (7.3%) were used in an algorithm. (Table I). Among low-magnification terms, most frequently used were ‘meshwork pattern’ (n=16), ‘ringed pattern’ (n=15), and ‘clod pattern’ (n=10) (Table II).

We grouped terms on the basis of similarity in definitions and found 40 ‘likely-synonymous’ clusters. By further excluding 12 terms lacking a clear definition, the total number terms could be reduced from 209 to 40 (80.8% reduction) (Table I).

High magnification / cellular-level resolution terms

Among 191 high-magnification terms, 182 terms had an identifiable definition and could be grouped into 32 ‘likely-synonymous’ clusters. Nine additional terms (6.6%) lacked a clear definition and were excluded from the likely-synonymous’ clustering and from further analysis. When stratifying the 182 individual high-magnification terms by their anatomic distribution, 68 terms mapped to the epidermis (37.4%), 46 terms to DEJ (23.9%), and 68 terms to the dermis (32.8%). When stratifying high-magnification terms by ‘likely-synonymous groups’, 13 terms (40.6%) mapped to the epidermis, 11 (34.3%) to the DEJ, and 8 (25.0%) to the superficial dermis.

Of 182 terms, 52 (28.5%) were simple terms and 130 (71.4%) were composite terms. We extracted 125 basic terminology units, 57 labeling the ‘cell / structure’ type, 51 describing the morphological attributes of these cells, and 17 relating to their anatomic distribution. We found that the 55 cells / structures could be categorized into eight recurring themes, namely basic terminology units describing: melanocytes as individual cells, dendritic processes without visible cell body, aggregated melanocytes, keratinocytes, DEJ structures, inflammatory cells, collagen structures and blood vessels (Table III).

Table 3:

Breakdown of reflectance confocal microscopy terms for melanocytic neoplasms into basic terminology units. Abbreviations: DEJ = dermo epidermal junction.

Cells / structures	Morphology modifiers/Descriptors	Anatomic location /Distribution
Relates to melanocytes as individual cells
Atypical cells	Dendritic	Pagetoid spread
Atypical melanocytes	Enlarged	Pagetoid pattern
Bright nucleated cells	Hyporeflective/dark	Pagetoid infiltration
Cell atypia	Irregularly bright	Widespread
Cells (NOS**)	Large	Focal
Cellular atypia	Pleomorphic	Folliculotropism
Cytological atypia	Round	Infiltrating adnexal structures
Melanocytic cells	Spindled	Non-specified
Melanocytes	With halo	In the dermal papillae
Nucleated cells		At the DEJ
Pagetoid cells		Dermal infiltration
		Infiltrating papillary dermis
		At the periphery
Relates to dendritic processes without visible cell body
Branching dendrites *	Composite dendrites
Dendritic processes *	Simple dendrites
Tangled lines *	Coarse dendrites
Relates to aggregated melanocytes
Aggregates of atypical cells	Cerebriform	Around the hair follicle
Bulging	Dense	Junctional
Cells clusters	Dense and sparse	Dermal
Clods	Dense dishomogeneous
Clusters	Dense irregular
Junctional thickening	Discohesive
Medusa head-like	Dishomogeneous
Melanocytic cell nests	Irregular
Mitochondria-like structures	Marked pleomorphism (within nest)
Nesting	Non-homogeneous
Nests	Pleomorphism (within nest)
Sheet-like	Sparse
Related to keratinocytes
Acanthosis	Atypical
Basal cells	Broadened (Honeycomb)
Cobblestone pattern	Irregular
Epidermal disarray	Monomorphic
Epidermal pattern	Poorly-defined
Honeycomb pattern	Readily detected
Keratinocyte cell borders	Regular
Refractile cells	Typical
Relates to DEJ structures/shape
DEJ disarray	Absent
Dermal papillae	Edged
Papillae	Loss of (papilla / papillary contour)
Papillary contours	Non-edged
	Non-visible
	Poorly visualized
	Ringed
Relates to stromal structures
Inflammatory cells
Bright dots	Round-to-triangular	Dermal
Bright non-nucleated cells
Bright particles
Bright small cells
Bright spots
Dermal bright cells
Inflammatory infiltrate
Inflammation
Melanophages
Plump cells
Stromal fibers
Bright fibrillar structures	Broadened
Collagen	Bundled
Curled fibers	Coarse
Fibroplasia	Reticulated
Highly reflecting fibers	Thickened
Stromal fibre
Thick cordons
Blood vessels
Vascularity	Atypical
Vessels	Enlarged
	Horizontal
	Irregular
	Prominent
	Tortuous morphology

^*Dendritic structures is a descriptive term that can pertain to dendritic processes emanating from either melanocytes or from Langerhans cells

^**NOS – not otherwise specified; the context of the original term implies that “cells” refer to melanocytes.

^***‘Grainy image’ and ‘epidermal granularity’ are not included in the table.

Low magnification / architectural patterns terms

By categorizing low-magnification terms into ‘likely-synonymous’ clusters, we reduced 18 individual terms to 8 ‘likely-synonymous’ group terms (55.5% reduction) (Table II). Three additional terms (16.6%) lacked a clear definition. When stratifying by anatomic level, all terms were ascribed to the DEJ and/or superficial dermis level and none to the epidermis.

DISCUSSION

In the present systematic review, describing the RCM terminology of melanocytic lesions, we extracted 209 terms from 59 studies. We grouped the individual RCM terms into ‘likely-synonymous’ clusters, based on similarity in definition and/or histopathology correlates, and excluded terms lacking a clear definition, potentially distilling 40 non-redundant terms from the 209 (80.8% reduction).

In a systematic review of RCM terminology for describing NMLs, we found that the average number of uses per term was only 1.6 times. Herein, we found that RCM terms from melanocytic neoplasms were used with an average frequency of 3.1. A higher average use frequency for melanocytic neoplasms, compared to NMLs, might be attributed to the previous consensus on melanocytic terminology in 2007.¹⁸ We also found in the present study a higher average use frequency for low-magnification terms, compared to high-magnification terms (average use of 3.4 vs 3.0, respectively). This may suggest higher consistency in the use of low-magnification RCM terms for melanocytic lesions.

We also found that RCM melanocytic terms are often composite (71.4%). This complexity may emanate from the fact that multiple RCM attributes need to be weighed to differentiate melanoma from challenging nevi – such as the shape, size and anatomic distribution of individual or aggregated melanocytes. Towards simplifying RCM terminology, we extracted 124 ‘basic terminology units’ from the 191 ‘high-magnification’ RCM terms, and categorized them into individual units describing ‘cells’ or ‘structures’ type, followed by the morphological attributes of these cells/structures and their anatomic distribution. This is akin to extracting basic words from complex sentences. This lays the groundwork for an expert agreement study that will eliminate redundancy and identify from the list the best basic descriptors (words) and composite criteria (sentences).

Creating a concise list of pertinent RCM terms for the diagnosis of melanocytic lesions is an important step towards widespread adaptation RCM. Along similar lines, Pellacani et al. recently published an expert consensus study, in which the experts identified 18 principal RCM terms for diagnosis of melanocytic and non-melanocytic lesions. These terms were then further clustered into two melanoma-specific key features (atypical cells (at any level) and DEJ disarray), one basal cell carcinoma-specific key features (basaloid cords/islands), and one squamous cell carcinoma-specific key feature (keratinocyte disarray). Identification of one of these four key features by novices was associated with a skin cancer diagnostic sensitivity of 91% and specificity of 57%.²¹ This study highlighted that a simplified shortlist of skin cancer ‘key-criteria’ could be easily learned and successfully utilized by novice RCM readers. A similar approach for triaging lesions has been proposed by simplified dermoscopy algorithms.⁷⁹. However, the RCM diagnosis of challenging melanocytic cases by experts requires a more elaborate set of criteria.

Limitations:

We included only full-text articles in English to allow direct comparisons of terms without translation bias. RCM terms published in non-English papers may have been missed. Non-peer reviewed articles (e.g. book chapters) were excluded and some terms used by the experts could have been missed. The weightage attributed to RCM terms may be influenced by the frequency of publications of individual researchers or research groups. Also, some terms were grouped together but further studies are needed to determine if more granularity is relevant or needed (e.g. collagen and elastosis). Finally, literature search was conducted on August, 2018, thereby, some recent key terms might have been overlooked.

Conclusion:

We propose a more concise glossary of RCM terms for the diagnosis of melanocytic lesions. By grouping the RCM terms, based on ‘likely-synonymous’ definitions, and by eliminating terms lacking a clear definition, the list could be reduced by over 80%. This systematic review lays the ground for an expert Delphi consensus for melanocytic lesions terminology. Furthermore, a concise glossary of RCM terms can facilitate standardizing of RCM diagnosis reports, teaching of novices, and communicating of research findings.

Capsule summary:

• Reflectance confocal microscopy (RCM) melanocytic neoplasms terminology is inconsistently used. Via systematic review, we identified redundant terms, and categorized 209 synonymous terms into 40 groups (80.8% reduction).

• The proposed shortened list of RCM-terms for melanocytic neoplasms may be easier to teach to novices and provides the basis for subsequent terminology consensus.