Abstract
About this talk
Primary dermal melanoma (PDM), an uncommon, non-metastatic melanoma with exclusive dermal or subcutaneous involvement, without epidermal component,1-3 has histological high-risk features of cutaneous metastases but with clinically-favorable outcome.1-4
Data of large, extended follow-up cohorts with detailed clinicopathologic characterization about PDM are limited. Hence, we evaluated clinical presentation, histopathologic features, immunohistochemical (IHC), genomic analysis (GA), American Joint Committee on Cancer/Tumor-Node-Metastasis (AJCC/TNM) staging, and clinical outcomes in the largest PDM cohort reported to date.
This IRB-approved, single-institution, retrospective-cohort study with PDM patients, diagnosed between 2000-2025, evaluated clinical and pathological data by using Epic (Epic Systems Corporation, Verona, WI, USA) and SoftPath (Softpath System, LLC, Norcross, GA, USA), respectively. Descriptive and comparative statistics were performed by using RStudio (Posit Software, PBC, Boston, MA, USA), with exact tests for categorical variables and summary measures for continuous variables.
Sixty-three patients (64 PDM cases), predominantly males (38/63, 60.3%) with a mean age of 63.1 years (SD 18.5), were identified. Lesions involved the head and neck (22%), trunk (15%), and upper extremities (10%), with nonrandom laterality distributions (all site-specific comparisons p<0.001). In cases with collected data, Breslow thickness (23/64, 35.9%) presented a mean of 2.93mm (SD 2.46) and mitotic activity (21/25, 84.0%), of 2.96/mm² (SD 2.56). Data for ulceration (4/17, 23.5%), lymphovascular invasion (14/14, 100%), microsatellitosis (13/13, 100%), neurotropism/perineural invasion (38/43, 88.4%), and regression (38/45, 84.4%) demonstrated high aggressive-like histopathological features in PDM.
IHC was performed in 47/64 cases (73.4%), mostly for HMB-45 (deep and scattered), S100, Melan-A, MART-1, and PRAME, with high positivity (12/18, 66.7%; 18/18, 100%; 32/36, 88.9%; 7/9, 77.8%; and 6/8, 75.0%; respectively). Less ordered, with variable positivity: AE1/AE3 (1/12, 7.7%), BAP1 (4/4, 100%), CD31 (5/6, 83.3%), CD34 (3/3, 100%), CD163 (1/1, 100%), CK5/6 (0/6, 0%), EMA (2/2, 100%), ERG (1/1, 100%), Ki-67 (mitotic activity>5%) (5/5, 100%), MITF (3/3, 100%), p16 (3/4, 75.0%), SOX10 (6/13, 46.2%), and tyrosinase (4/5, 80.0%). These data show the diagnostic complexity of PDM and the importance of clinicopathologic correlation.
AJCC/TNM staging data were available for 39/63 cases (61.9%). AJCC stages I, II, III, and IV accounted for 30.8%, 23.1%, 30.8%, and 15.4%, respectively, without significant difference between early (I/II) and advanced (III/IV) disease (p=0.491). Advanced primary tumor classification (T3-T4) was observed in 19/39 cases (48.7%), indifferent from T1-T2 tumors (p=1.000). Regional-nodal and distant metastasis was present in 15/40 (37.5%; p=0.718) and 6/39 cases (15.4%; p=1.000), respectively. Median follow-up was 6.39 years (IQR 1.65-8.98). Melanoma-specific mortality was low (1/34 patients; 2.9%).
Limited GA was performed in five cases, revealing characteristics commonly seen in melanoma, including 9p loss (2/5, 40%) and 7q and 8q gains, supporting the absence of a distinct molecular signature in PDM.³
These findings reinforce prior observations that PDM exhibits distinct biologic behavior from, favorable survival than, and poor-prognosis features like melanoma.1-5 (Supplemental: Table 1, Figure 1) Compared to previously-published small series,1 this cohort’s data support recognition of PDM as a distinct entity. Global consortium, incorporating standardized staging, complete outcome ascertainment, and integrated molecular profiling, are needed to further clarify prognostic determinants and guide management. 3,5
Speaker
About the speaker
