Background: While adjuvant therapies are currently recommended for patients with stage IIB-IV melanoma, biomarkers are missing to improve risk stratification to select the most efficient treatment and patients to be included in clinical trials. We genetically characterized a large cohort of patients with stage-II-melanoma to identify predictive and prognostic biomarkers.
Methods: Clinical data of 193 stage-II-melanoma patients from the German Central Malignant Melanoma Registry were identified. All patients were therapy-naïve at the time of primary resection and received no adjuvant treatment until recurrence. Tumour-normal pairs were sequenced with a comprehensive cancer panel.
Results: 30.1% of the tumours were classified as BRAF-mutated, 28.0% as RAS-, 18.1% as NF1-, and 23.8% as Triple-WT. In-silico prediction identified a potential new candidate driver, CBL, in 10.4% of the patients. GISTIC nominated deletions of region 11q23.1-3 containing CBL as a potential driver alteration. Enrichment of mutations in CBL was replicated in published cohorts. Patients with RAS-mutated tumours and CBL deletions showed worse OS (p = 0.004) and RFS (p = 0.044). Point mutations in CBL were highly enriched in patients with NF1-mutated tumours, showing a trend (p = 0.18) towards worse OS.
Conclusions: Our findings suggest CBL as a novel driver gene in melanoma, which is mutated in a relevant fraction of patients. Deletions of 11q23.1-3 including CBL were identified as a prognostic marker indicating a higher risk of recurrence and shorter survival. Furthermore, CBL could support patient stratification to identify high-risk patients who may benefit from adjuvant therapies or intensified monitoring strategies.
© 2026. The Author(s).