Melanoma with peripheral globules: clinical and dermoscopic features
Morphology and biology of nevi with peripheral globules are well known, while that of melanomas with peripheral globules remain unclear. We analyzed lesions with peripheral globules from 10-90-years old and identified features beyond the peripheral globules to distinguish benign from malignance. These observations can reduce unnecessary biopsies. A total of 401 melanocytic lesions with peripheral globules were included in this retrospective study. Dermoscopic patterns and structures, including those of the peripheral globules, were evaluated. Of the 401 lesions, 179 (44.64%) were excised, 41 (10.22%) of which were melanoma. Melanomas were most common in the lower extremities (P <0.01), with a disorganized pattern, whereas melanocytic nevi were most common on the trunk, with an organized pattern. The presence of blotches, atypical dots and globules, or atypical vessels was associated with melanomas (P <0.01). We concluded melanocytic lesions displaying peripheral globules are at the greatest risk for melanoma if located on the lower extremity and lesions reveal any of the following structures:blotch or atypical dots and globules or atypical vessels. In this Mendeley data we show examples of melanocytic nevi and melanomas with peripheral globules, the flow chart, the management algorithm proposed, supplementarey tables and the graphical abstract.
Steps to reproduce
This longitudinal, observational, and retrospective study was approved by the ethics committee of the A.C. Camargo Cancer Center. All clinical notes made between 2006 and 2020 from medical records at a dermatology tertiary cancer center were selected using the following descriptions: “enlarging nevi,” “growth nevi,” “peripheral globules,” and “melanoma with peripheral globules.” Clinical and dermoscopic documentation was performed using a FotoFinder Dermatoscope dermoscopy device. Inclusion criteria for each lesion were as follows: melanocytic lesion presenting dermoscopy with PGs in at least 30% of the circumference; in-focus polarized dermoscopy images; data available in medical records with the lesion outcome such as the biopsy report if the lesion was excised or digital dermoscopic images taken at least one year after the baseline imaging for lesions deemed to be benign and subjected to follow-up. The dermoscopic analysis was performed by an experienced dermoscopist (A.F.A.M.) blinded to the diagnosis. After this, two other experts (J.C.T.B.; T.C.B.M.P.) reviewed until a consensus was reached. The following features for each lesion was described: clinical features of melanocytic lesions; dermoscopic pattern features (globular, reticular, reticulo-globular, homogeneous, and multicomponent); dermoscopic peripheral globules features and dermoscopic melanocytic lesion global features. All statistical analyses were performed with IBM SPSS statistics software version 23.0, and R version 3.5.0. A descriptive analysis of the data was presented considering the patient’s sample unit and the lesion. To assess the possible factors that lead to cutaneous melanoma characterization, a generalized estimating equation (GEE) model with binomial distribution for the response variable and logit link function was fitted to the dataset. This model allows estimation of correlation among lesions in the same patient, once several patients present two or more lesions. In every fitted model, the interchangeable correlation (correlation among patients’ lesions was the same) was considered. Melanocytic nevi and melanomas with peripheral globules were compared regarding clinical, morphological, and dermoscopic features. Based on these findings, a management algorithm was created.