The role of central nervous system (CNS) prophylaxis with high-dose methotrexate (HDMTX) in DLBCL is controversial. In this retrospective study, we evaluated the efficacy of prophylactic HDMTX on isolated CNS relapse, concomitant CNS and systemic relapse, systemic relapse, and survival outcomes in 226 patients with newly diagnosed DLBCL and high-risk CNS International Prognostic Index (CNS-IPI) score treated with RCHOP. The three-year risk of isolated CNS relapse was significantly lower in patients who received HDMTX, at 3.1% compared to 14.6% (P = 0.032) in those who did not. However, neither concomitant CNS-systemic relapse rates, systemic relapse rates, nor three-year PFS and OS were significantly different between treatment groups in multivariable analysis. Among propensity score-matched patients (N = 102), HDMTX was also associated with significantly lower isolated CNS relapse rates (HR 0.06, 95% CI 0.004–0.946, P = 0.046). HDMTX was well tolerated with manageable toxicities when given at a dose of 3 g/m² by day 3 of RCHOP chemotherapy. Using propensity score matching and multivariable regression to yield treatment groups with well-balanced covariates, we showed that prophylactic HDMTX improved isolated CNS relapse rates but did not decrease concomitant CNS-systemic relapse rates, systemic relapse rates, or improve survival outcomes.

使用大剂量甲氨蝶呤 (HDMTX) 预防中枢神经系统 (CNS) 在 DLBCL 中的作用存在争议。在这项回顾性研究中，我们评估了预防性 HDMTX 对 226 名新诊断的 DLBCL 和高危 CNS 国际预后指数 (CNS-IPI) 评分的患者的孤立性 CNS 复发、伴随 CNS 和全身性复发、全身性复发和生存结局的疗效用 RCHOP 治疗。接受 HDMTX 的患者 3 年单纯 CNS 复发的风险显着降低，为 3.1%，而未接受的患者为 14.6% ( P  = 0.032)。然而，在多变量分析中，治疗组之间伴随的 CNS 全身复发率、全身复发率以及 3 年 PFS 和 OS 均无显着差异。在倾向评分匹配的患者中（N  = 102），HDMTX 还与显着降低孤立的 CNS 复发率相关（HR 0.06, 95% CI 0.004–0.946, P  = 0.046）。在RCHOP 化疗的第 3 天以 3 g/m ²的剂量给药时，HDMTX 耐受性良好，毒性可控。使用倾向评分匹配和多变量回归来产生具有均衡协变量的治疗组，我们发现预防性 HDMTX 提高了孤立的 CNS 复发率，但没有降低伴随的 CNS 全身复发率、全身复发率或改善生存结果。