Identifying a novel ferroptosis-related prognostic score for predicting prognosis in chronic lymphocytic leukemia

BackgroundChronic lymphocytic leukemia (CLL) is the most common leukemia in the western world. Although the treatment landscape for CLL is rapidly evolving, there are still some patients who develop drug resistance or disease refractory. Ferroptosis is a type of lipid peroxidation-induced cell death and has been suggested to have prognostic value in several cancers. Our research aims to build a prognostic model to improve risk stratification in CLL patients and facilitate more accurate assessment for clinical management. MethodsThe differentially expressed ferroptosis-related genes (FRGs) in CLL were filtered through univariate Cox regression analysis based on public databases. Least absolute shrinkage and selection operator (LASSO) Cox algorithms were performed to construct a prognostic risk model. CIBERSORT and single-sample gene set enrichment analysis (ssGSEA) were performed to estimate the immune infiltration score and immune-related pathways. A total of 36 CLL patients in our center were enrolled in this study as a validation cohort. Moreover, a nomogram model was established to predict the prognosis. ResultsA total of 15 differentially expressed FRGs with prognostic significance were screened out. After minimizing the potential risk of overfitting, we constructed a novel ferroptosis-related prognostic score (FPS) model with nine FRGs (AKR1C3, BECN1, CAV1, CDKN2A, CXCL2, JDP2, SIRT1, SLC1A5, and SP1) and stratified patients into low- and high-risk groups. Kaplan-Meier analysis showed that patients with high FPS had worse overall survival (OS) (P<0.0001) and treatment-free survival (TFS) (P<0.0001). ROC curves evaluated the prognostic prediction ability of the FPS model. Additionally, the immune cell types and immune-related pathways were correlated with the risk scores in CLL patients. In the validation cohort, the results confirmed that the high-risk group was related to worse OS (P<0.0001), progress-free survival (PFS) (P=0.0140), and TFS (P=0.0072). In the multivariate analysis, only FPS (P=0.011) and CLL-IPI (P=0.010) were independent risk indicators for OS. Furthermore, we established a nomogram including FPS and CLL-IPI that could strongly and reliably predict individual prognosis. ConclusionA novel FPS model can be used in CLL for prognostic prediction. The model index may also facilitate the development of new clinical ferroptosis-targeted therapies in patients with CLL.