1325PValidation of a clinicopathological and gene expression profile (CP-GEP) model for sentinel lymph node metastasis in primary cutaneous melanoma

Abstract Background As primary cutaneous melanoma patients with a positive sentinel lymph node (SLN, stage III) are now considered candidates for adjuvant systemic therapy, a SLN biopsy (SLNB) is indicated in more patients. However, SLNB is an invasive procedure and is negative in approximately 80% of patients. Therefore, there is a need for a non-invasive test to accurately identify patients with primary cutaneous melanoma without nodal metastases. Here we describe the first independent validation of a recently developed CP-GEP model to predict nodal metastasis. This risk model combines Breslow thickness, age, and gene expression variables from the primary melanoma. Methods This study focused on all patients >18 years who underwent a SLNB at the Erasmus Medical Center (between January 2007 and December 2017), within 90 days after diagnosis of primary cutaneous melanoma. Total RNA was extracted from formalin-fixed paraffin-embedded (FFPE) primary cutaneous melanomas, reversed transcribed into cDNA and subsequently analyzed for the expression of 8 target genes involved in melanoma metastasis (ITGB3, PLAT, SERPINE2, GDF15, TGFBR1, LOXL4, IL8, MLANA) using an optimized qPCR protocol. Results FFPE tissue samples from 211 patients were analyzed using the CP-GEP model. At diagnosis, the median age was 55 years (interquartile range [IQR] 45-65), and the median Breslow thickness was 2.1mm (IQR 1.4-3.4). Most patients presented with a T2 or T3 melanoma, accounting for 94 and 70 patients, respectively. Overall, 27.5% of patients had a positive SLN. The CP-GEP model had a negative predictive value (NPV) of 89.4%. In patients with stage T1-T2 melanoma, the model was able to achieve an SLNB reduction rate of 40.1% with an NPV of 90.7%. Conclusions The CP-GEP model is a non-invasive and validated tool that is able to predict nodal metastasis in an independent Dutch population. Consequently, this risk model is able to accurately identify patients with primary cutaneous melanoma that can safely forego SLNB. Therefore, the CP-GEP model is a promising tool for patient care, preventing unnecessary surgery in the majority of patients. Legal entity responsible for the study Erasmus University Medical Center, Department of Dermatology. Funding SkylineDx. Disclosure J.T. Dwarkasing: Shareholder / Stockholder / Stock options, Full / Part-time employment: SkylineDx BV. D. Tempel: Shareholder / Stockholder / Stock options, Full / Part-time employment: SkylineDx BV. L. Bosman: Shareholder / Stockholder / Stock options, Full / Part-time employment: SkylineDx. A.A.M. Van der Veldt: Advisory / Consultancy: BMS; Advisory / Consultancy: MSD; Advisory / Consultancy: Roche; Advisory / Consultancy: Novartis; Advisory / Consultancy: Pierre Fabre; Advisory / Consultancy: Pfizer; Advisory / Consultancy: Sanofi; Advisory / Consultancy: Ipsen. All other authors have declared no conflicts of interest.