Performance of a 31-gene expression profile (GEP) test for metastatic risk prediction in cutaneous melanomas (CM) of the head and neck.

9576 Background: Accurate prognostication of distant metastatic risk using sentinel lymph node (SLN) biopsy for CM can be challenging in melanomas of the head and neck due to a higher false negative rate compared to other anatomical areas. A GEP signature that predicts metastatic risk based on primary tumor biology, providing a binary outcome of Class 1 (low risk of metastasis) or Class 2 (high risk), was previously described. The prognostic capabilities of the GEP independently and in combination with SLN status in a cohort of patients with primary head and neck CM are assessed here. Methods: All samples and clinical data were collected under an IRB-approved multicenter protocol. qPCR analysis was used to assess expression of the gene signature (Class 1 vs. Class 2). Distant metastasis-free survival (DMFS) and melanoma-specific survival (MSS) were assessed. Results: 157 subjects with primary CMs in the head and neck region were identified. 110 of 157 subjects had a SLN biopsy performed. Median age was 65 years (range 25-89) and median Breslow depth was 1.6 mm (range 0.2-15.0 mm). In 71 SLN-negative patients, 18 of 27 (67%) distant metastatic events were GEP Class 2. Overall, 73% (47 of 64) distant metastases, and 88% (22 of 25) deaths due to CM were called Class 2. By comparison, sensitivities for DMFS and MSS were 41% (26 of 64) and 52% (13 of 25), respectively, using SLN biopsy alone, and increased to 80% (51 of 64) and 88% (22 of 25), respectively, when combining the SLN status and GEP class. Kaplan-Meier 5-year DMFS and MSS rates based on SLN status alone or in combination with GEP are shown in the table. Conclusions: These data support the ability of the GEP test to accurately identify low- and high-risk cases of head and neck melanoma. The results strongly support the role of GEP testing to enhance current staging by better predicting the risk of distant metastasis and death for patients with melanoma in an anatomic region that is associated with a higher SLN biopsy false negative rate. [Table: see text]