Background: The prevalence and cancer-specific death rate of lung cancer (LC) have risen in recent decades. A universally applicable prognostic signature for both adenocarcinoma LC (LUAD) and squamous cell carcinoma LC (LUSC) is still lacking.
Methods: A total of 453 patients from The Cancer Genome Atlas (TCGA)-LUAD cohort and 452 patients from TCGA-LUSC cohort were enrolled, and a prognostic model was constructed using least absolute shrinkage and selection operator (LASSO) regression analysis based on the consensus prognostic genes in both cohorts. The newly defined pan-lung cancer risk count (PLCRC) of each patient was calculated via the summation formula.
Results: A total of 23 genes were selected for the calculation of the PLCRC. The PLCRC showed a moderate prognostic value in the entire (p < 0.001, HR: 2.75, AUC: 0.643), LUAD (p < 0.001, HR: 2.51, AUC: 0.636) and LUSC (p < 0.001, HR: 2.89, AUC: 0.656) cohorts. The PLCRC was an independent prognostic factor after adjusting the clinical features. The PLCRC was also effective in nine external validation cohorts and in patients with different clinical features. Activation of extracellular matrix pathways and infiltration of immunocytes promoted the tumorigenesis and development of both LUAD and LUSC. We generated a universally applicable prognostic signature, the PLCRC, which could dichotomize patients with significantly different clinical outcomes and guide the clinical treatment of LC patients. Chemotherapy is more suitable for patients with a low PLCRC, while anti-cytotoxic T-lymphocyte-associated protein 4 immunotherapy is more suitable for patients with a high PLCRC.
Conclusion: We established and validated a newly defined prognostic signature, the PLCRC, for both LUAD and LUSC patients and provided clinical strategies for patients from different risk subgroups.
Keywords: LASSO analysis; immunotherapy; lung cancer; personalized treatment; prognostic signature.
© 2021 Chen et al.