Advancing melanoma treatment: personalized medicine through molecular targeting. Construction and analysis of a Comprehensive clinical database and identification of biomarkers for adjuvant therapy in stage III melanoma

Melanoma, arising from melanocytes, constitutes a small percentage (5%) of skin cancers but is the leading cause of mortality in this category due to its high metastatic potential. Stage III melanoma, characterized by regional lymph node involvement without distant metastases, represents a pivotal phase with a significant risk of progression. While advancements in adjuvant therapies, including targeted molecular therapy and immunotherapy, have significantly improved survival outcomes, particularly progression-free survival (PFS), disease recurrence and progression remain substantial challenges. The clinical course in stage III is highly heterogeneous, highlighting the limitations of relying solely on AJCC staging and underscoring the urgent need for reliable prognostic and predictive biomarkers to guide individualized therapeutic approaches. To investigate these challenges, the research involved three complementary approaches: a retrospective observational study of 218 stage III melanoma patients treated at the AOU Città della Salute e della Scienza di Torino (2016-2024); a translational study on circulating tumor DNA (ctDNA) in 35 patients on adjuvant therapy, which included the development of a dynamic risk model; and a pilot Next- Generation Sequencing (NGS) study on 8 melanoma samples. A comprehensive database integrating clinical, demographic, and molecular data was constructed on the RedCap platform. Key findings confirmed that disease stage and the presence of ulceration are significant prognostic factors. The BRAF mutation was associated with a higher likelihood of progression. In this cohort, targeted therapy (Dabrafenib + Trametinib) showed a significantly better PFS compared to immunotherapy (p=0.001). In the ctDNA study, while baseline levels showed only a weak trend towards significance, a novel dynamic risk model based on baseline levels and their early changes showed a potential tool in risk stratification. The pilot NGS study identified mutations in genes such as LRP1B, BARD1, and PTEN and revealed a notable discrepancy in BRAF mutational status in two patients compared to routine testing, highlighting the importance of advanced technologies. In conclusion, while adjuvant therapies have improved outcomes, managing advanced melanoma remains challenging. The creation of an integrated database is a fundamental resource for future research. ctDNA shows promise as a dynamic biomarker for monitoring treatment response and detecting early relapse. Although preliminary, these findings strongly support the integration of multiple biomarker types for more precise and personalized melanoma management, pointing to the need for larger, multicentric studies to validate these results and further improve patient outcomes.