The efficacy of high-dose versus intrathecal methotrexate for DLBCL with high risk of CNS relapse: A multicenter retrospective study

High-dose methotrexate (HD-MTX) has been used as central nervous system (CNS) prophylaxis in high risk DLBCL, but its efficacy to prevent CNS relapse is controversial. Here we aimed to evaluate the prophylactic effect of HD-MTX on CNS relapse by direct comparison with intrathecal methotrexate (IT-MTX) in high risk DLBCL. This multicenter retrospective study was conducted at six institutions. Patients diagnosed with DLBCL at a high risk of CNS relapse between 2003 and 2020 were treated with frontline chemotherapy. The risk of CNS relapse was defined: (1) involvement of the specific extranodal sites (testis, breast, adrenal gland, kidney, bone, or paranasal sinus), and (2) CD5-positive DLBCL. Patients received CNS prophylaxis after achieving an overall response to initial chemotherapy with IT-MTX (2003-2013) and HD-MTX (2014-2020). Among 132 eligible patients (98 given HD-MTX and 34 given IT-MTX). Median age was 62.5 years (22-84), 34.1% had two or more extranodal site involvements, 23.5% had a high CNS-IPI score and 31.1% were positive for CD5. The median follow-up for the entire cohort was 52 months (9-174), and 11 patients had isolated CNS relapse: 6 (6.1%) in the HD-MTX group and 5 (14.7%) in the IT-MTX group. All CNS relapses involved brain parenchymal lesions. The median time until CNS relapse was 38 months (HD-MTX 11-56, IT-MTX 13-122). The incidence of CNS relapse at 3 years was 3.9% in the HD-MTX group and 6.1% in the IT-MTX group (P = 0.93). Using propensity score-matched analysis, there was no significant difference in the risk of CNS relapse between the prophylaxis routes (4.6% HD-MTX vs. 7.6% IT-MTX, P = 0.86). We found no difference in CNS relapse between the HD-MTX and IT-MTX populations. The results were similar after adjusting for background factors with a propensity score-matched analysis. However, a larger number of patients and longer follow-up periods are needed to confirm this conclusion.