In 2–8% of patients with melanoma, the first clinical manifestation of the disease may be skin metastasis. In these cases, differential diagnosis with the primary melanoma, benign melanocytic lesions, and other malignant and benign skin growths is particularly challenging. For this reason, the dermatologist's approach to cutaneous metastases of malignant melanoma calls for knowledge of the great morphological variety of these lesions. Dermoscopic characteristics associated with CMMMs have not yet been codified. The aim of the present review is to provide additional information about dermoscopic aspects of these skin lesions.
Cutaneous malignant melanoma is the most aggressive and metastatic form of skin cancer with unpredictable pathways of spread. Although it comprises <5% of malignant skin tumors, it is responsible for almost 60% of lethal skin cancers. In Europe, national cancer registries show a rising incidence (3–8% per year) of melanoma in the past two decades, higher in females than males. Mortality rate is high, surpassed only by lung cancer. An estimated 16.5% of patients with primary melanoma develop in-transit, regional, or distant recurrences. The clinical and histological features of primary lesions are not predictive of metastatic patterns. Cutaneous melanoma may metastasize to different organs with variable frequency: skin and lymph nodes (42–57%); lungs (18–36%); liver (14–20%); brain (12–20%); bone marrow (11–17%); and small intestine (1–7%). Thus, nearly half of patients with metastatic melanoma will have cutaneous metastases. The latter are a frequent event in the natural history of malignant melanoma, both in the early and late phases of disease progression. In fact, melanoma is the second in frequency only to breast cancer in the development of skin metastases. In 2–8% of patients with melanoma, the first clinical manifestation of the disease may be the skin metastasis. In these cases, differential diagnosis with the primary melanoma, benign melanocytic lesions, and other malignant and benign skin growths is particularly challenging.
The dermatologist's approach to cutaneous metastases of malignant melanoma (CMMMs) calls for knowledge of the great morphological, clinical, and dermoscopic variety of these lesions. Metastatic melanoma may mimic benign lesions, and their differentiation is critical, as prognosis and choice of therapy in relation to stage are completely different.
Dermoscopy is a non-invasive diagnostic technique in routine use by dermatologists, especially in Europe.[6, 7] The many conferences and courses organized in the last few years have produced criteria for distinguishing melanomas from benign pigmented skin lesions with increasing accuracy, creating somewhat complex but effective new semiotics of pigmented skin lesions. The conventional dermoscopic analysis of skin lesions is based on recognition of specific criteria and application of different diagnostic algorithms.[8-12] Much data are available on the various dermoscopic patterns of skin tumors, but there are few data on patterns of CMMMs.[13-15] For this reason, it seemed appropriate to prepare this update on the various dermoscopic patterns shown by CMMMs.
Global Dermoscopic Patterns
Cutaneous metastases of malignant melanoma do not have distinctive dermoscopic characters but rather certain patterns that can be found with varying frequency in other benign and malignant skin lesions.[14-16]
This has uniform diffuse pigmentation (red, brown, gray, or grayish black) without any other definite structures (Fig. 1a). In particular, the brown homogeneous pattern sometimes can mimic a nevus of Clark (Fig. 2). A similar pattern may be observed in blue nevi (Fig. 1b), basal cell carcinoma, melanomas (Fig. 1c), and hemangiomas (Fig. 1d).[17-20] Indeed, blue nevi show a global pattern characterized by homogeneous bluish or steel blue pigmentation without any pigment network or other distinctive dermoscopic structure. In most cases, the blue color is combined with brown, gray, or black. Instead, non-pigmented and lightly pigmented variants of basal cell carcinoma frequently show a shiny white to red background (about 83%), and heavily pigmented basal cell carcinoma shows a multicolored background (blue, red, light brown, gray, dark brown, or black) in about 67% of cases. Hemangiomas have a homogeneous reddish background, while melanoma shows diffuse areas of hyper- or hypopigmentation.[19, 20] Thus this global dermoscopic feature needs to be combined with specific local features to obtain indications for diagnosis.
This results from oval junctional nests containing atypical melanocytic cells (Fig. 3a). Colorimetric variations of this pattern depend on the amount of neovascularization and melanic pigment, produced by atypical cells that proliferate without observing anatomical limits and therefore invade various layers of the epidermis in a disorderly way. The resulting combinations are mainly bluish red, pale brownish red, brownish, or grayish red and blue–gray. Although this is a highly specific pattern for malignancy (about 99%), it is also often found in blue nevi (Fig. 3b), basal cell carcinomas (Fig. 3c), Clark nevus (Fig. 3d), and hemangiomas. Indeed, hemangiomas have given rise to confusion because their hematic lacunae may be very similar to metastatic saccules, both being uniform in color and structure with sharp borders.
This is reported in about 32% of CMMMs (Fig. 4a), much less than in common nevi, primary melanoma (Fig. 4b), and blue nevi (Fig. 4c). Amelanotic lesions are not clinically striking due to their dermal localization and lack of pigment. A more detailed history, palpation, and dermoscopy analysis are useful for diagnosis. The most frequent vascular morphologies seen under dermoscopy were respectively serpentine vessels, corkscrew-like vessels, arborizing vessels, dotted vessels, and milky-red areas. Moreover, dermoscopic evaluation of amelanotic cutaneous melanoma metastases also revealed angioma-like lacunae, poorly demarcated, without septae of division but with hairpin or serpentine vessels in the lacunae-like areas. Crystalline structures may be another clue for the recognition of this type of metastases.
This is significantly the most frequent in CMMMs (53%) and shows characteristics related to tumor thickness: punctate vessels predominate in thin lesions (Fig. 5) and corkscrew vessels in thick ones (Fig. 6). The latter have an irregular linear spiral form and are typical of melanoma metastases (about 83%), though they can also be found in amelanotic areas of primary nodular tumors (melanomas). Vascular structures/areas are certainly more common in CMMMs than in primary cutaneous melanomas, as accentuated neoangiogenesis is a typical factor of neoplastic progression and metastatic dissemination. The particular disposition of vascular structures could be a valid diagnostic indicator, making it possible to differentiate metastatic and primary melanoma. In the former case, they are observed close to the tumor edges, whereas in the latter case they are mainly central.[22, 23] However, also for this pattern, differential diagnosis cannot ignore blue nevi, which mainly contain arborizing vessels, and only in a small percentage of cases (about 12%) can vascular structures such as polymorphic, punctate, comma, and irregular linear be observed. Differential diagnosis cannot ignore hemangiomas, in which the tumultuous proliferation of blood vessels show red and bluish lacunar areas determined by dilated blood vessels in the superficial dermis, crossed by whitish strands. Likewise it cannot ignore melanoma, typically variegated, showing punctate, irregularly distributed hairpin, linear, and irregular vessels or polymorphic vessels in areas of regression.[8, 21, 23]
The distribution of dermoscopic structures and colors within the lesion is chaotic. This pattern is observable in many pigmented and nonpigmented skin lesions (Fig. 7a–c).
Focal Dermoscopic Patterns
After evaluation of overall characteristics, dermoscopic analysis should examine certain specific structures. Color is an important aspect of clinical approach as it enables diagnosis to be narrowed down.
Reddish color (bluish or brownish gray)
This may suggest primitive melanoma and CMMM (Fig. 8a) by virtue of tumor neoangiogenesis, which implies increased fragility of blood vessels due to the many atypical morphological forms. It may also suggest angiomatoid lesions with typical vascular dilations in the superficial dermis, which may develop thromboses. Gray–brown coloring is mainly observed in primitive melanomas and benign tumors and can therefore be considered a negative indicator of CMMM.
Peripheral gray spots/globules
These are due to fine melanin particles or melanin “dust” in melanophages or free in the deep papillary or reticular dermis. They can be observed in about 25% of CMMMs, where they are irregularly distributed at the lesion periphery (Fig. 8b). They may be confluent and thus resemble globules. They have different appearances and distribution and can be associated with different background colors. Grayish spots are seen mainly in high-grade dysplastic nevi, in situ, and early invasive melanomas (about 93% of melanomas), as well as benign lesions (about 26%). Their localization (focal/diffuse), distribution (center/off-center), morphological appearance (regular/irregular), and association with other colors (red/blue–gray/white) and hypopigmentation (yes/no) are other important features.
Perilesional erythema (polymorphic vessels)
This is observed in very few cases of CMMM but is highly specific for melanomas and blue nevi (Fig. 8c).
Pigmentary peripheral halo
In about 40% of CMMMs, pigmentary peripheral halo is another important feature (Fig. 8d) but can also be observed in common nevi, basal cell carcinoma, and melanomas. In this last case, the peripheral halo usually is not symmetrically distributed around the lesion.
Dermoscopy is an essential tool for dermatologists who encounter and treat skin cancers. This technique makes it possible to observe and recognize certain morphological features, peculiar to various skin lesions and difficult or impossible to evaluate with the naked eye. This facilitates diagnosis in general, but what is more important, it enables early detection of certain structural characteristics of early malignant melanoma. Indeed, various articles on the topic show that when expert dermatologists use dermoscopy based on pattern analysis, the diagnostic accuracy of pigmented skin lesions may improve by 10–27% with respect to diagnosis based on naked eye examination. Many research groups have formulated morphological diagnostic criteria based on specific patterns, colors, intensity of pigmentation, configuration, regularity, margins, surface features, and so forth.
Dermoscopic characteristics associated with CMMMs have not yet been codified, but certain patterns, although almost never pathognomonic, are rather typical. The broadest and most recent study describing dermoscopic features of skin secondaries of metastatic melanoma is the statistical analysis by Bono et al. This study considered nine dermoscopic characters chosen on the basis of statistical significance and specificity in the first genuine review by Schulz in 2000. Bono and coworkers recognized the greater prevalence of dermoscopic characters such as vascular and saccular patterns, light brown halo, and peripheral erythema in CMMMs, identifying the vascular pattern and pigmented halo as strong indicators of CMMM.[15, 27]
It would be very important for the specialist who deals with this type of injury to know the correlation/specificity of the different dermoscopic patterns found in CMMMs. Currently, in the literature, there are few data in this regard because the observed CMMMs do not possess the necessary numerical significance to be classified into categories. However, we can provide the reader with some guidance based on what was observed in our series. We interpreted the data relating to the percentage frequency of nine patterns in our lesions. Based on our observations, we could identify three categories, defined as follows: strong (when a pattern is found in a percentage of cases between 66 and 100%) moderate (if observed between 33 and 66% of CMMMs); and weak (<33% of lesions). None of the analyzed patterns falls into the first category (neither in our series, nor in that of Bono et al.). A moderate correlation emerges, in our cases, for the homogeneous pattern and the blue color; instead in the Bono et al.' series, for the homogeneous and vascular pattern, pigmentary peripheral halo, and brown–gray color. The third category is the most crowded, demonstrating weak specificity of most of the patterns described, due to the low amount of lesions analyzed up to now in the literature. In fact, in our cases, this category includes the saccular, vascular, amelanotic, and polymorphic pattern, peripheral gray spots and globules, perilesional erythema, pigmentary peripheral halo, and pink–red, brown–gray, and black color. A similar situation occurs in the Bono et al.' series with the saccular, amelanotic, and polymorphic pattern, peripheral gray spots, perilesional erythema, and pink–red, blue, and black color.
Our analysis, inspired to dermoscopic patterns described by Bono et al., has revealed some differences, not only, as already said above, in terms of percentage frequency but also in appearance; we refer mainly to the peripheral halo. Bono et al. suggested that “pigmentary” was the peripheral halo observed in its CMMMs (in 40.8% of lesions); on the contrary, our halo was in most cases whitish (15.92%; Fig. 9); a pigmentary halo was seen in only 3.53%. This discrepancy is probably due to some structures being nuanced and highly dependent on the observer and type of instrument used. Indeed, the whitish halo, probably representing fibrosis, is seen much more clearly under a polarized light dermoscope.
The aim of the present update on CMMMs was to provide additional information about dermoscopic aspects of these lesions as some studies demonstrate that early detection of CMMMs may substantially improve patient survival. Although the goals of treatment of CMMMs are generally palliative, with emphasis on improvement of function, cosmesis, and prevention of complications, such as infection, ulceration, and local destruction, much hope has been invested in new treatments such as cell therapy, gene therapy, and targeted therapy, where we see the great medical, scientific, and legal interest of the question.[30, 31] As it concerns new drugs for advanced disease, it underlines that the new agents specifically targeted against melanoma are changing the standard of care for this disease. The monoclonal antibody ipilimumab is the first new agent approved for first- and second-line treatment of metastatic melanoma, based on improved overall survival compared with standard therapy.[31, 32] Other drugs, such as BRAF inhibitors, have demonstrated high activity in advanced disease. Researchers are already planning ways to combine some of these new drugs to render a possible longer-term control of metastatic tumors. These new drugs may also change the natural history of metastatic disease according to the stage when we are able to make the correct diagnosis. This also forces us to remember the importance of medicolegal aspects related to melanoma. In fact, as it concerns legal considerations, it is to remind that they can arise in any aspect of dermatological practice but are more likely in the diagnosis and treatment of skin cancers, especially melanoma. Recent papers showed that errors in diagnoses comprise the largest average indemnity and, in particular, the second greatest number of paid claims for dermatologists. Thus, errors in diagnosis and failure to perform (diagnostic or therapeutic procedures) account for the greatest risk. Melanoma is the most common diagnostic error. It is also the most costly error. Thus, the primary steps that a practicing dermatologist can take to minimize risk are to follow accepted guidelines and the indicated procedures in coming to a diagnosis.
In conclusion, it would be useful to conduct multicentric studies on large populations to determine more clearly dermoscopic characteristics of melanoma metastasis. A future aim is to go beyond interpretative variability linked to observer experience and to create algorithms enabling easier framing of these skin lesions that are often initially difficult to interpret. However, histological examination remains fundamental. Dermoscopy can be a valid diagnostic support, helping to differentiate simulators and to recognize aspects important for prognosis, therapy, and legal purposes at an early stage.