ICE therapy for primary and secondary central nervous system involvement of lymphomas

Abstract

Abstract Central nervous system (CNS) lymphomas, including primary CNS lymphoma (PCNSL) and secondary CNS lymphoma (sCNSL), are aggressive and challenging diseases with dismal outcomes. Due to its CNS-penetrating properties, ifosfamide, carboplatin, and etoposide (ICE) protocol with or without rituximab has become a frequently used salvage regimen for these lymphomas. However, the short-term efficacy and toxicity of this therapeutic approach as well as its effect on subsequent treatment approaches such as autologous stem cell transplantation (ASCT) and chimeric antigen receptor T-cell (CAR-T) therapy, remain limited. This study aimed to evaluate clinical outcomes following R-ICE therapy in patients with CNS lymphoma. A retrospective analysis of consecutive adult patients with primary or secondary CNS lymphoma treated with R-ICE at a large tertiary center. Patient demographics, disease characteristics, treatment response, subsequent therapy with ASCT/CAR-T, toxicity, and overall survival (OS) were analyzed. Twenty-four patients were included with a median age of 71 years (range 33-85), 11 of them females. The majority of patients had newly-diagnosed isolated PCNS lymphoma (62.5%), while the remainder had DLBCL with CNS involvement at the initial diagnosis (12.5%), relapsed/refractory DLBCL (20.8%), and mantle cell lymphoma (MCL, 4.2%). The median number of R-ICE cycles was 2, with 87.5% receiving between one and three cycles. ICE dose reductions were common, with a median dose intensity of 90%. Approximately half of the patients received a reduced ICE dose (<100%). 37.5% of patients proceeded to autologous stem cell transplantation (ASCT), 16.7% received CAR-T cell therapy, and 4.2% underwent allogeneic transplantation. No further treatment was given to 41.7% of patients. Among all patients, 29% achieved a complete metabolic response (CMR), 33% had a partial response (PR), and 37.5% failed to achieve response. At a median follow-up of 11 months (range: 0–58 months), 75.0% (18/24) of patients had died. A trend toward better 2y-OS was observed among patients who achieved a CMR (2y-OS 57%) compared to those without CMR responses (2y-OS of 17%) (χ²(1) = 2.08, p = 0.150). All CMR patients (100%) proceeded to advanced therapies, including ASCT or CAR-T therapy, whereas only 41% of non-CMR patients transitioned to such treatments (χ²(1) = 4.85, p = 0.028). Treatment-related toxicity was tolerable. 6 patients (25.0%) demonstrated grade>3 non-hematologic-toxicity (encephalopathy, nausea, cystitis). Early-death (within 30 and 60 dats of ICE initiation) was documented in 35% and 40% of patients, respectively. Prior methotrexate (MTX) exposure did not appear to influence treatment response. The leading cause of death was disease progression, followed by sepsis and one case of CAR-T associated ICANS. In summary, R-ICE regimen is effective and tolerable in patients with CNS involvement and can serve as a bridge to advanced and potentially curative approaches. These findings support the role of R-ICE in CNS lymphoma and highlight areas for further prospective investigation.