Background
Open-label, randomized phase 3 trial (56 centers, 5 European countries); 368 enrolled, 230 randomized (229 analyzed); newly diagnosed, immunocompetent B-cell primary CNS lymphoma (PCNSL), age 18–70; all received 4 cycles of MATRix induction (rituximab, high-dose methotrexate, high-dose cytarabine, thiotepa); patients reaching ≥partial response randomized 1:1 to two consolidation strategies.
Results
Interventions and follow up: Arm A: High-dose chemotherapy + autologous stem-cell transplantation (HCT-ASCT); carmustine + thiotepa conditioning (n=114)
Arm B: 2 cycles R-DeVIC — rituximab, dexamethasone, etoposide, ifosfamide, carboplatin (non-myeloablative) (n=115)
Primary endpoint: Progression-free survival (PFS)
mFollow up: 45.3 mo
Results: PFS (3-yr): 78% (HCT-ASCT) vs 51% (R-DeVIC); HR 0.43 (95% CI 0.27–0.68); P=.0003
OS: significantly superior with HCT-ASCT vs R-DeVIC (both PFS and OS improved; HCT-ASCT established as preferred consolidation)
Arm B: 2 cycles R-DeVIC — rituximab, dexamethasone, etoposide, ifosfamide, carboplatin (non-myeloablative) (n=115)
Primary endpoint: Progression-free survival (PFS)
mFollow up: 45.3 mo
Results: PFS (3-yr): 78% (HCT-ASCT) vs 51% (R-DeVIC); HR 0.43 (95% CI 0.27–0.68); P=.0003
OS: significantly superior with HCT-ASCT vs R-DeVIC (both PFS and OS improved; HCT-ASCT established as preferred consolidation)
Adverse events
Mean AEs per patient: 9.3 (R-DeVIC) vs 14.6 (HCT-ASCT)
Fatal serious AEs (post-consolidation): R-DeVIC 2 (both secondary AML); HCT-ASCT 5 (infections n=4, pulmonary embolism n=1)
Overall: more toxicity with HCT-ASCT, but treatment-related mortality low in both arms
Fatal serious AEs (post-consolidation): R-DeVIC 2 (both secondary AML); HCT-ASCT 5 (infections n=4, pulmonary embolism n=1)
Overall: more toxicity with HCT-ASCT, but treatment-related mortality low in both arms
Conclusions
In the largest randomized trial in untreated PCNSL to date, thiotepa-based HCT-ASCT significantly improved progression-free and overall survival versus non-myeloablative R-DeVIC consolidation after uniform MATRix induction. HCT-ASCT is established as the preferred consolidation strategy in transplant-fit patients who respond to induction.
Key Limitations
Only patients achieving at least a partial response to MATRix induction were randomized (roughly one-third of enrolled patients were never randomized), selecting for chemosensitive disease and limiting generalizability to induction non-responders. Open-label design. R-DeVIC is a less-established comparator than whole-brain radiotherapy or other non-transplant regimens used in some regions. HCT-ASCT carried higher treatment-related mortality (5 fatal serious AEs). Eligibility was restricted to fit patients ≤70 years, and long-term neurocognitive and quality-of-life outcomes were not the primary readout.
Clinical Context
Consolidation after high-dose methotrexate-based induction in PCNSL has historically pitted whole-brain radiotherapy (effective but neurotoxic) against myeloablative HCT-ASCT or non-myeloablative chemoimmunotherapy. MATRix/IELSG43 provides the first large randomized head-to-head demonstrating HCT-ASCT superiority over non-myeloablative consolidation, reinforcing NCCN, ESMO, and EHA preference for thiotepa-based HCT-ASCT consolidation in transplant-eligible, induction-responsive PCNSL.