A Novel Prognostic Model for Primary CNS Lymphoma Incorporating Clinico-Laboratory Parameters

Abstract Background This study aimed to identify prognostic factors in patients with newly diagnosed primary central nervous system lymphoma (PCNSL) treated with high-dose methotrexate–based therapy and to develop a novel risk-stratification model using easily measurable clinical and laboratory parameters. Methods A total of 451 patients with newly diagnosed PCNSL were identified from a prospective registry at Asan Medical Center, Seoul. Patients were randomly assigned to a training cohort (n = 280; October 2002–August 2019) and an independent validation cohort (n = 171; September 2019–December 2023). Results With a median follow-up of 106.0 months (95% CI, 101.0–120.0), the median overall survival (OS) in the training cohort was 46.1 months (95% CI, 34.9–57.6). Independent predictors of worse OS (p < 0.05) included age ≥65 years, high serum β2-microglobulin levels (≥1.8 mg/L), elevated serum lactate dehydrogenase, and ECOG performance status >1. These four factors were combined to form the ABLE score, which stratified patients into low- (0 risk factors), intermediate- (1 risk factor), and high-risk (≥2 risk factors) groups. In the training cohort, median OS was 109.0, 49.0, and 18.0 months, respectively (p < 0.001). Validation in the independent cohort confirmed significant prognostic discrimination, with median OS of not reached, 53.1, and 19.0 months for each risk group (p < 0.001). Comparative analyses demonstrated that the ABLE model showed improved discrimination compared with existing systems. Bootstrap validation (n = 451) yielded an optimism-corrected C-index of 0.656 (95% CI, 0.628–0.685). Conclusions The ABLE risk-stratification model can effectively differentiate prognostic subgroups in patients with PCNSL.