Computational recognition of LncRNA signatures in tumor-associated neutrophils could have implications for immunotherapy and prognostic outcome of non-small cell lung cancer.

Cancer immune function and tumor microenvironment are governed by long noncoding RNAs (lncRNAs). Nevertheless, it has yet to be established whether lncRNAs play a role in tumor-associated neutrophils (TANs). Here, a computing framework based on machine learning was used to identify neutrophil-specific lncRNA with prognostic significance in squamous cell carcinoma and lung adenocarcinoma using univariate Cox regression to comprehensively analyze immune, lncRNA, and clinical characteristics. The risk score was determined using LASSO Cox regression analysis. Meanwhile, we named this risk score as "TANlncSig." TANlncSig was able to distinguish between better and worse survival outcomes in various patient datasets independently of other clinical variables. Functional assessment of TANlncSig showed it is a marker of myeloid cell infiltration into tumor infiltration and myeloid cells directly or indirectly inhibit the anti-tumor immune response by secreting cytokines, expressing immunosuppressive receptors, and altering metabolic processes. Our findings highlighted the value of TANlncSig in TME as a marker of immune cell infiltration and showed the values of lncRNAs as indicators of immunotherapy.