Abstract
Objective: To identify and validate prognostic protein biomarkers for primary central nervous system lymphoma (PCNSL) and develop a novel prognostic scoring system integrating protein biomarkers with clinical characteristics.
Methods: This study employed a four-pronged methodological approach:
1. Retrospective high-throughput DIA proteomic analysis of cerebrospinal fluid (CSF) from PCNSL patients with favorable vs unfavorable prognoses to identify differential proteins.
2. Prospective PRM-targeted proteomic validation using pretreatment CSF from newly diagnosed PCNSL patients.
3. ELISA validation of key biomarkers and IHC assessment of their expression in brain tissues.
4. Development of a prognostic scoring system through univariate/multivariate analysis of clinical characteristics and protein biomarkers, with subsequent risk stratification and survival analysis.
Results: Firstly,we identified 395 differentially expressed proteins, with 9 (IGLV4-69, AGRN, XPO1, HPR, FGFR3, PRELP, LFNG, CGREF1, CRTAC1) validated by PRM. XPO1 showed the strongest prognostic discrimination (P<0.0001; ROC-AUC=0.824). Secondly,multivariate analysis revealed ECOG >2, elevated XPO1, and elevated IGLV4-69 as independent poor prognostic factors. Thirdly, ELISA and IHC assays consistently showed higher XPO1 expression in poor-prognosis patient. Fourthly,we developed a prognostic scoring system incorporating six parameters (1 point each): multiple lesions, bulky disease, ECOG>2, elevated CSF protein, CSF lymphoma cells, and high level of XPO1. Patients were stratified into low-intermediate risk (0-3 points) and high-risk (4-6 points) groups, with significantly different outcomes: median PFS 23 vs 2 months (HR=3.878, 95%CI:2.137-7.039, P<0.001) and median OS not reached vs 19 months (HR=4.606, 95%CI:2.187-9.701, P<0.001).
Conclusion: Through integrated proteomic and clinical analysis, we identified XPO1 as a key PCNSL prognostic biomarker and developed a practical scoring system that effectively discriminates risk groups. The established scoring system may guide personalized treatment strategies and facilitate targeted therapeutic decisions of PCNSL patients.
