Genome-wide screening and immune landscape suggest a potential-m6A-related lncRNA risk signature for predicting prognosis of melanoma.

BACKGROUND

Melanoma is the most dangerous form of skin cancer because of its high metastatic potential. Potential-N6-methyladenosine (m6A)-related long noncoding RNAs (pMRlncRNAs) play a vital role in malignancy. The identification of prognostic-related pMRlncRNAs and development of risk signatures could improve the prognosis and promote the precise treatment of melanoma.

METHODS

Gene expression and relevant clinical data were obtained from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) databases. Prognostic-related pMRlncRNAs were selected using univariate Cox regression analysis. Patients with melanoma were classified into different subtypes using the "ConsensusClusterPlus" package, and the ESTIMATE algorithm was applied to depict their immune landscape. A pMRlncRNA risk signature was developed using least absolute shrinkage and selection operator regression analysis and verified using survival analysis and receiver operating characteristic curves. Gene set enrichment analysis (GSEA) was used to investigate the underlying biological pathways. The relationships between risk score and clinicopathological characteristics, as well as programmed cell death-ligand 1 (PD-L1) expression level, were investigated. A nomogram with calibration curves was established to comprehensively predict the outcome of melanoma.

RESULTS

Fifteen pMRlncRNAs were significantly associated with overall survival (OS). Two cluster subtypes were identified by consensus clustering. Patients in cluster 2 were associated with better OS, higher PD-L1 expression level, lower T stage, and higher ESTIMATEScore, ImmuneScore, and StromalScore than those in cluster 1. There were differences in immune cell infiltration between the 2 clusters. Ten pMRlncRNAs with prognostic value were selected to develop a risk signature, that functioned as an independent prognostic factor for melanoma. Patients with low-risk scores had a better prognosis in general. The area under the curve (AUC) value (0.720), as well as 1-, 3-, and 5-year calibration curves, revealed that the risk signature has suitable predictive power for prognosis. GSEA revealed 10 pathways that might play important roles in melanoma. Moreover, patients with high-risk scores were associated with advanced T stage, cluster 1, lower ImmuneScore, and higher PD-L1 expression level.

CONCLUSIONS

We developed a novel 10-pMRlncRNA risk signature that could elucidate the crucial role of pMRlncRNAs in the immune landscape of melanoma and predict prognosis.