Dermatoscopic features of thin (≤ 2mm Breslow thickness) versus thick (> 2mm Breslow thickness) nodular melanoma and predictors of nodular melanoma versus nodular non‐melanoma tumors: a multicentric collaborative study by the International Dermoscopy Society

2020 
BACKGROUND Thin nodular melanoma (NM) often lacks conspicuous melanoma-specific dermatoscopic criteria, and escapes clinical detection until it progresses to a thicker and more advanced tumor. OBJECTIVE To investigate the dermatoscopic morphology of thin (≤2mm Breslow thickness) versus thick (>2mm) NM and to identify dermatoscopic predictors of its differential diagnosis from other nodular tumors. METHODS Retrospective, morphological case-control study, conducted on behalf of the International Dermoscopy Society. Dermatoscopic images of NM and other nodular tumors from 19 skin cancer centers worldwide were collected and analyzed. RESULTS Overall, 254 tumors were collected (69 NM of Breslow thickness ≤ 2 mm, 96 NM >2mm and 89 non-melanoma nodular lesions). Light brown coloration (50.7%) and irregular brown dots/globules (42.0%) were most frequently observed in ≤2mm NMs. Multivariate analysis revealed that dotted vessels (3.4-fold), white shiny streaks (2.9-fold) and irregular blue structureless area (2.4-fold) were predictors for thinner NM compared to non-melanoma nodular tumors. Overall, irregular blue structureless area (3.4-fold), dotted vessels (4.6-fold) and serpentine vessels (1.9-fold) were predictors of all NM compared to non-melanoma nodular lesions. LIMITATIONS Absence of a centralized, consensus pathology review and cases selected form tertiary centers maybe not reflecting the broader community. CONCLUSIONS Our study sheds light into the dermatoscopic morphology of thin NM in comparison to thicker NM and could provide useful clues for its differential diagnosis from other non-melanoma nodular tumors.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    40
    References
    5
    Citations
    NaN
    KQI
    []