PCNSL Patients Treated with the Matrix Regimen in the Real World Have Similar Outcome to Those in the IELSG 32 Trial

Introduction Identifying optimal treatment for primary CNS lymphoma (PCNSL) patients is essential to improve their poor survival rates. The addition of rituximab and thiotepa to a methotrexate/cytarabine-based regimen (i.e., MATRix), followed by consolidation therapy, was shown to be effective in the phase II IELSG 32 trial. The endpoints after induction and consolidation treatment in the IELSG 32 trial were complete remission (CR) rate and progression-free survival (PFS). We aimed to compare outcomes of PCNSL patients treated with the MATRix regimen who met the key inclusion criteria (IC) of the IELSG 32 trial in the real world (RW) with outcomes in that trial (first presented at ICML, Lugano, 2015). Methods Consecutive patients with PCNSL were prospectively collected in the NiHiL registry (NCT03199066) between 2015-2022 (n = 290), including 280 individuals receiving systemic chemotherapy (CHT), and 88 (31%) patients treated with the MATRix regimen. Of them, 78 (89%) met key IC of the IELSG 32 trial (age 18-70 years, PS ECOG 0-3 and 0-2 for 66-70 years) and were included in the analysis (MATRix-RW). The remaining 10 patients were excluded (age > 70 years in 8 cases, PS ECOG 4 in one case, PS ECOG 0-3 with age 66-70 years in one case). Among patients who met the key IC of the IELSG 32 trial, the proportion of patients treated with MATRix regimen increased over time from 24% in 2015 to 56% in 2022. Responses to therapy and survival of MATRix-RW group were compared with the MATRix arm of the IELSG 32 trial (MATRix-trial). Results Median age at diagnosis was 57 years for both MATRix-RW and MATRix-trial populations. PS ECOG > 1 at diagnosis was observed in 46% vs 32% of the MATRix-RW vs MATRix-trial patients. A high-risk IELSG index was observed in 10% vs 20% of the MATRix-RW vs MATRix-trial patients. Induction CHT. The median number of MATRix cycles in the MATRix-RW group was 4 (range 1-4). In 7 (9%) MATRix-RW patients the MATRix regimen was substituted by less intensive CHT during induction due to infections or toxicity (information in the MATRix-trial). A total of 6 (8%) vs 3 (4%) patients died in MATRix-RW vs MATRIX-trial groups during induction CHT. Following induction CHT, 67 MATRix-IC patients had responsive or stable disease (CR 58%, PR 24%, SD 4%) vs 66 MATRix-trial patients (CR 49%, PR 37%, SD 1%), and were screened for consolidation eligibility. Consolidation. A total of 52 vs 51 MATRix-RW vs MATRix-trial patients (out of CR/PR/SD pts) were eligible for consolidation. Reasons for ineligibility included poor mobilization of peripheral stem cells (n = 2 vs 4 among MATRix-RW vs MATRix-trial patients), early progression (n = 2 vs 7), unperformed leukapheresis (n = 1 among MATRix-trial group), and other (n = 11 vs 3). Among consolidation-eligible patients, two MATRix-RW vs three MATRix-trial patients refused further therapy. Consolidation plan (autologous stem cell transplant, ASCT, vs whole brain radiotherapy, WBRT) in the MATRix-RW population was based on the treating center algorithm vs randomization in the MATRix-trial. ASCT was administered in 37 (47%) vs 25 (33%) MATRix-RW vs MATRIX-trial patients, and WBRT in 13 (17%) vs 23 (31%) patients. The CR rate of MATRix-RW vs MATRix-trial patients treated with ASCT was 84% vs 93%, and with WBRT 85% vs 95%. Survival. Median follow-up of the MATRix-RW vs MATRix-trial patients was 52 vs 88 months. The PFS of the MATRix-RW vs MATRix-trial patients was 58% vs 61% at 2 years, and 47% vs 52% at 7 years; the OS was 61% vs 69% at 2 years, and 51% vs 56% at 7 years. The PFS of the MATRix-RW patients consolidated with ASCT and WBRT was 69% and 84% at 7 years. The OS of the MATRix-RW vs MATRix-trial patients consolidated with ASCT was 74% vs 70% at 7 years, and with WBRT 84% vs 71% at 7 years. Notably, patients treated with MATRix who didn't meet IC for IELSG 32 trial (n = 10) had significantly worse survival compared to the MATRix-RW (2-year OS 30% vs 61%, HR 2.30, P = 0.043). Conclusions PCNSL patients who meet key IC for the IELSG 32 trial and receive the MATRix regimen in the real world have similar survival to those treated in the MATRix arm of the trial. No significant survival differences were observed among patients consolidated with ASCT and WBRT. With a larger ASCT-consolidated population of MATRix-treated PCNSL patients in comparison to the IELSG 32 trial, our analysis confirms the benefit of MATRix/ASCT treatment approach in the real world. Supported by: Charles Uni Hem-Onco Cooperatio Program, grant NU21-03-00411.