Descriptive study reveals prevalence and characteristics of high-risk neoplasias in skin samples, suggesting need for better assessments.
INTRODUCTION: Turnaround time (TAT) for routine skin specimens may be prolonged at times, delaying diagnosis and treatment in cases with a unexpected high-risk neoplasia. This quality improvement study assessed the prevalence, types and characteristics of such cases. METHODS: This was a descriptive study of all routine skin specimens submitted by general practitioners (GPs) and private dermatologists in 2022-2023 at the Department of Pathology, Viborg Regional Hospital, Denmark. From Systematized Nomenclature of Medicine (SNOMED) codes, various premalignant and malignant neoplasias were identified; melanocytic and soft tissue neoplasms were included. RESULTS: Among 33,280 routine specimens, 77 (0.23%) were a high-risk premalignant or malignant neoplasia: 66 melanocytic (54 melanomas, 0.16%) and 11 sarcomas. Superficial spreading melanoma (n = 36) and the in situ variant (n = 10) were the most frequent high-risk neoplasia in lesions clinically assessed as benign. Among melanocytic lesions, 61% had involved/unassessable margins, and 67% of melanomas had a depth ≥ 1.0 mm. The mean TAT was 64 calendar days. Mean age: 64.1 years, with a notable proportion aged 41-50 years. Referrals: GPs 73%, dermatologists 26%, surgeons 1%. PROCEDURES: excision 58%, curettage 25%, biopsy 10%, unspecified 7%. Suspicious shape/colour features were noted in 36% of melanocytic lesions. 7.5% had prior melanoma. CONCLUSIONS: Unexpected high-risk neoplasia are rare (0.23%) but often advanced, highlighting the need for improved clinical assessment and a short TAT for routine skin specimens. The results serve as a reference for future AI-based screening. FUNDING: None. TRIAL REGISTRATION: Not relevant.
No takes yet. Share an insight, caveat, or question.
Egsgaard et al. (2026) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: