One of the main prognostic factors for melanoma is the thickness of the primary tumor, measured in millimeters from the granular layer of the epidermis or base of a superficial ulceration to the deepest location of the malignant cells, excluding follicular invasion and with an approximation of 0.1mm.1,2 The greater the thickness, the greater the risk of metastasis.3 For a thickness <1.0mm, in the absence of ulceration and a number of mitoses >1permm2, the sentinel lymph node positivity rate is less than 6%, being at least 14% for a thickness >1.0mm. Therefore, in the latter case, the sentinel lymph node biopsy is recommended.4 In addition to the decision on sentinel lymph node screening, the thickness also influences the extension of the resection margin.3,5 The biopsy of a lesion suspected of melanoma should be excisional, whenever possible, so that the therapeutic decision can be based on precise staging parameters.2 Incisional biopsies can underestimate thickness in 5% to 22% of cases,6 based on sampling that does not always represent the area of ​​greatest tumor thickness.

Dermoscopy is a technique that allows increased diagnostic accuracy in melanoma.7 Some studies have shown its capacity to infer histopathological characteristics, namely thickness, mitotic index, and the presence of ulceration.8–12 However, the most recent literature is not abundant in this context and the concrete application of the results has been little explored.

Thus, the primary objective of this study was to assess the relationship between melanoma thickness and dermoscopic variables and, secondly, between tumor thickness and demographic, epidemiological, and clinical variables, seeking to identify predictive factors of melanoma thickness.

An analytical, observational, and cross-sectional study was carried out on a sample of patients diagnosed with cutaneous melanoma, at the Dermatology and Venereology Service of Centro Hospitalar e Universitário de Coimbra, between January 2014 and December 2017.

The sample selection considered the following inclusion criteria: definitive diagnosis of cutaneous melanoma located on the head and neck, trunk, upper limb, hand, lower limb, or foot and the presence of clinical data related to the dermoscopic and anatomopathological examinations.

The clinical records of the selected patients were consulted to collect demographic (age and sex), epidemiological (phototype, eye color, hair color at 18 years of age, presence of ephelides, presence of solar lentigines, history of sunburn with erythema and blisters, number of melanocytic nevi >2mm in size, according to the definition of the International Agency for Research on Cancer, presence of atypical nevi according to the ABCDE criteria and presence of actinic keratoses), clinical (melanoma location, asymmetry, border, color, change in dimension in the last 12 months and change in color in the last 12 months), dermoscopic (dermoscopy pattern, presence of atypical pigment network or pseudo-network, presence of negative network, presence of irregular dots or globules, presence of hypopigmentation, presence of striae or pseudopods, presence of blue-gray veil, presence of atypical vessels and presence of white shiny lines) and anatomopathological (Breslow thickness) data.

The statistical analysis was performed using the SPSS® software, version 25, with a significance level set at 0.05. The variables were represented through their relative and absolute frequency. The chi-square test and Fisher’s test were used to establish associations between the collected variables. Univariate and multivariate binary logistic regression allowed establishing the predictive factors of thickness >1mm.

The study was carried out according to the recommendations of the Declaration of Helsinki of the World Medical Association and was approved by the Ethics Committee of Centro Hospitalar e Universitário de Coimbra.

This study did not receive any specific funding from public, private or non-profit agencies.

A total of 314 patients with cutaneous malignant melanoma were diagnosed in the aforementioned time period. One hundred and ninety-five cases were excluded due to a lack of recording of the dermoscopic variables being assessed, and the remaining 119 cases were included.

The demographic and clinical characteristics of the individuals included in the study are shown in Table 1. Most patients (94.1%; n=112) were at least 40 years old at the time of the diagnosis and were male (51.3%; n=61). Malignant melanomas were located mainly on the trunk (35.3%; n=42). Similar frequencies were recorded for melanoma thickness.

The multicomponent dermoscopic pattern (54.2%) was associated with a thickness >1mm, while the reticular pattern (52.5%) was the most frequent in melanomas with a thickness <1mm (p<0.001). The presence of an atypical pigment network/pseudo-network (86.4%; p<0.001) and negative pigment network (8.6%; p=0.027) were associated with a thickness ≤1mm, while the presence of an atypical vascular pattern (65.0%; p<0.001) and white shiny lines (18.3%; p=0.003) was more frequent in melanomas with a thickness >1mm (Table 2).

Table 3 compares the distribution of demographic and epidemiological variables among patients with malignant melanoma with different thicknesses. The frequency of individuals over 40 years old (98.3%; p=0.049) was significantly higher when the thickness was >1mm and this characteristic was associated with a lower frequency of ephelides (1.7%; p=0.032) and previous sunburns (42.1%; p=0.032). Sex, phototype, eye color, hair color, presence of solar lentigines, number of melanocytic nevi, presence of atypical nevi, and actinic keratoses did not differ significantly in relation to melanoma thickness.

The absence of pigmentation (amelanotic melanoma) was associated with thickness >1mm (17.5%; p<0.001). No significant differences were observed regarding melanoma location, asymmetry, border, color change, and recent growth in relation to thickness (Table 4).

Table 5 shows the results of the univariate and multivariate logistic regression on predictive factors of thickness >1mm in patients with malignant melanoma. In the univariate analysis, the presence of ephelides, previous sunburns, and atypical pigment network/pseudo-network were predictive of thickness <1mm, but the presence of an atypical vascular pattern or white shiny lines was related to thickness >1mm. The multivariate regression model explained 49.7% of the variance associated with the thickness (R2=0.497) and identified the atypical vascular pattern (relative risk of 6.621; p<0.001) as an independent predicitve factor of thickness >1mm. Instead, atypical pigment network/pseudo-network (relative risk 0.168; p=0.001) was identified as a predicitve factor of thickness <1mm.

Of the assessed clinical and epidemiological variables, only the presence of ephelides and a history of previous sunburns were identified as predictive factors for thin melanomas, facts not consistently reported in previous studies.13 This result can be explained by the acknowledgment of ephelides and sunburns as risk factors for the development of skin neoplasms and the consequent adoption of secondary prevention measures, such as the use of regular skin self-examination.

Previous studies have shown an association between the atypical pigment network/pseudo-network and thin melanomas, as well as an association between the atypical vascular pattern and white shiny lines with thicker melanomas.8,9 However, as opposed to other series, the presence of a blue-gray veil and striae was not associated with melanomas thicker than 1mm.8–10

After the multivariate regression, only an atypical vascular pattern and atypical pigment network/pseudo-network were highlighted respectively as independent predicitve factors of melanomas thicker than 1mm and melanomas with thickness ≤1mm or less. These associations suggest that more invasive melanomas have greater neovascularization, depicted at the dermoscopic examination as atypical vessels and that melanomas with a thickness ≤1mm maintain the dermal-epidermal junction intact since the atypical pigment network corresponds to an atypical junctional melanocytic proliferation.14

Some study limitations are highlighted, such as the lack of data on some of the assessed variables and the absence of an optimal correlation between the dermoscopic analysis, performed on a horizontal plane, and the anatomopathological examination, performed on the vertical plane. Moreover, in the future, it will be important to carry out prospective cohort studies to confirm the obtained results.

An incisional biopsy carries the risk of underestimating melanoma thickness, leading to the understaging and eventual inaccuracies in defining the appropriate excision margins and the indication or not of sentinel lymph node biopsy. Dermoscopy is a potentially useful technique for reducing understaging errors, allowing the choice of the most appropriate site for the incisional biopsy, which should avoid areas where there is only an atypical pigment network (related to inferior tumor thickness) and preferably focus on areas where atypical vessels are observed (associated with greater thickness).

None declared.

Ana Rita Silva: Statistical analysis; design and planning of the study; drafting and editing of the manuscript; collection, analysis, and interpretation of data; critical review of the literature.

Ricardo Vieira: Approval of the final version of the manuscript; design and planning of the study; drafting and editing of the manuscript; collection, analysis, and interpretation of data; effective participation in research orientation; critical review of the manuscript; intellectual participation in the propaedeutic and/or therapeutic conduct of the studied cases.

None declared.