From mechanistic synergy to clinical breakthrough: Updated results from a phase II Trial of HD-MTX, selinexor, and zanubrutinib in CNS lymphoma

Abstract

Abstract Background: Central nervous system lymphoma(CNSL) is an aggressive extranodal non-Hodgkin lymphoma, most commonly classified as diffuse large B-cell lymphoma (DLBCL). High-dose methotrexate (HD-MTX)-based regimens remain standard of care despite complete response rates of ~50%. Alternatively, Bruton's tyrosine kinase inhibitors (BTKis) suppress malignant B-cell proliferation by blocking FOXO3a degradation, but are also limited by acquired resistance and poor prognosis. Selinexor (KPT-330) is a first-in-class oral Exportin-1 inhibitor that effectively crosses the blood–brain barrier (BBB): it promotes FOXO3a nuclear retention and synergizes with BTKis to inhibit PI3K/AKT/NF-κB signaling. Given selinexor's potential to overcome BTKi resistance and improve outcomes beyond HD-MTX, we evaluated its combination with HD-MTX and BTKi in preclinical models and a phase II trial.Methods:To assess selinexor–BTKi synergy in vitro, we treated OCI-LY10 cells (ABC-type DLBCL) with fixed-ratio selinexor and BTKi based on IC₅₀ values. We assessed drug interactions by the Chou-Talalay method, measured apoptosis, and evaluated key signaling proteins including JAK/STAT3, NF-κB, NFATC1, NOTCH1, and BCL-2. We also treated A20 tumor-bearing mice with the combination treatment for 30 days, and then analyzed apoptosis (Ki67, BAX, BCL-XL), immune infiltration (CD3, CD68, CD163, PD-1), and T-cell subsets (CD4, CD8, TIM-3, PD-1) in tumor and spleen tissues.Building on these preclinical results and our 2023 ASH preliminary data (n=5, complete response rates [CRR] 100%), the phase II trial (ChiCTR2200062154) extended to 21 CNSL patients (13 males, 8 females) with a median age of 63 years (range: 18–80 years). These patients were treated with the MSZ regimen comprising six 21-day cycles of HD-MTX (3.5 g/m², day 1), selinexor (40 mg twice weekly), and zanubrutinib (160 mg twice daily). We quantified selinexor levels in plasma and cerebrospinal fluid (CSF) using LC-MS/MS at 2, 4, 6, 8, and 25 hours. Endpoints included ORR (primary endpoint), PFS (progression-free survival), and OS (overall survival), based on the Lugano 2014 criteria.Results:Selinexor and BTKi exhibited strong synergy in vitro (CI=0.41), suppressing proliferation and inducing apoptosis in a dose- and time-dependent manner. The combination blocked JAK/STAT3 and NFATC1 pathways, while NOTCH1 and BCL-2 showed compensatory upregulation. In the A20 mouse model , the combination reduced tumor volume by >80% (P<0.001), increased CD3⁺ T-cell infiltration 2.5-fold, decreased CD163⁺/CD68⁺ macrophages by 60%, downregulated PD-1 and TIM-3, and expanded CD4⁺/CD8⁺ T-cell populations in the spleen.In the clinical cohort, 20/21 patients (95.2%) achieved complete response after a median follow-up of 13.8 months. Sixteen (76.2%) experienced early, deep, durable remissions after two (n=9) and four (n=7) cycles. Updated survival outcomes showed 1-year PFS of 94.1% and OS of 100%. One patient underwent autologous stem cell transplantation, seven received radiotherapy, and 13 continued on selinexor–zanubrutinib maintenance (six of whom completed 2 years). Grade 3 hematological toxicity was limited to two patients (9.5%), and no treatment-related mortality was observed. Two patients died of unrelated causes. Pharmacokinetic analysis showed selinexor plasma levels peaked at 4 hours (average of 204.67 ng/mL) after a 40 mg oral dose, with corresponding CSF levels at 0.919 ng/mL. Serial sampling from 4 to 25 hours demonstrated a stable CSF-to-plasma ratio (~0.6%).Conclusions: Our findings confirm that selinexor, alone or with BTKi, exerts potent anti-tumor effects in ABC-type DLBCL by inhibiting proliferation, inducing apoptosis, inhibiting JAK/STAT3 and NFATC1 signaling, and remodeling the immune microenvironment through T-cell activation and macrophage reprogramming. Clinically, the MSZ regimen exhibits superior efficacy (CR 95.2%; 1-year PFS 95.2%; 1-year OS 100%) compared to current standard therapies for CNSL, with the majority of patients achieving rapid, deep, and durable remissions. This is also the first in-human quantification of selinexor's BBB penetration, with pharmacokinetic profiling establishing a novel “plasma–CSF” exposure-effect model. These data support individualized dosing and lay the foundation for overcoming BTKi resistance positioning selinexor–BTKi as a mechanistically grounded therapeutic backbone under clinical evaluation for CNSL.