A randomized phase III trial (MATRix/IELSG43) provides the first direct phase III evidence that high‑dose myeloablative chemotherapy followed by autologous haematopoietic stem‑cell transplantation (HCT–ASCT) is superior to a non‑myeloablative chemoimmunotherapy consolidation strategy for selected patients with primary central nervous system lymphoma (PCNSL). This research highlight summarizes the trial population, treatment pathways, primary end point and the reported conclusion that supports HCT–ASCT as a standard consolidation option in transplant‑eligible patients who respond to induction therapy.
The MATRix/IELSG43 trial enrolled a total of 346 patients with previously untreated B‑cell PCNSL who were considered eligible for transplantation. All enrolled patients initiated a planned induction course; only those who completed induction and achieved at least a partial response proceeded to the randomized consolidation phase. In total, 230 patients met those criteria and were randomized 1:1 to one of two consolidation strategies.
Induction therapy in the trial was the MATRix regimen, composed of high‑dose methotrexate and cytarabine combined with thiotepa and rituximab. This high‑dose methotrexate–based induction approach is a commonly used backbone in PCNSL treatment. After four cycles of MATRix, patients who achieved a partial response or better were eligible for randomization into the consolidation comparison.
Patients who completed induction with response were randomized to receive either a carmustine–thiotepa‑based HCT–ASCT consolidation or two cycles of a non‑myeloablative chemoimmunotherapy regimen, R‑DeVIC (rituximab plus dexamethasone, etoposide, ifosfamide and carboplatin). The trial therefore compared an intensive myeloablative transplant approach with a less intensive, non‑myeloablative consolidation strategy in a population selected for transplant fitness and treatment response.
The pre-specified primary end point of the randomized comparison was progression‑free survival measured from the time of randomization. According to the trial report highlighted in Nature Reviews Clinical Oncology, the MATRix/IELSG43 results confirm the superiority of HCT–ASCT over the non‑myeloablative R‑DeVIC consolidation in the randomized cohort of transplant‑eligible patients who had responded to induction.
Before MATRix/IELSG43, phase III evidence directly demonstrating benefit of myeloablative HCT–ASCT versus non‑myeloablative consolidation in PCNSL was lacking. The randomized data from this trial therefore support adoption of HCT–ASCT as a standard‑of‑care consolidation option following successful high‑dose methotrexate–based induction in appropriately selected patients.
Key caveats for clinical application include the trial’s selection criteria: patients were transplant‑eligible and had achieved at least a partial response to MATRix induction before randomization. These selection factors are essential when extrapolating the results to broader PCNSL populations, including older or medically unfit patients who may not meet transplant eligibility criteria.
The Nature Reviews Clinical Oncology highlight summarizes the trial’s design, population numbers and the principal conclusion that HCT–ASCT is superior in this setting. Detailed numerical results (for example, hazard ratios, absolute event rates, overall survival data, toxicity and quality‑of‑life outcomes) and subgroup analyses were not reported in the highlight itself. For complete methodological details, full outcome metrics and safety data, the original phase III report in The Lancet (Illerhaus et al., MATRix/IELSG43) should be consulted.
The primary randomized trial is cited as Illerhaus G. et al., High‑dose chemotherapy followed by autologous stem‑cell transplantation versus non‑myeloablative consolidation in primary CNS lymphoma (MATRix/IELSG43): a randomised phase 3 trial, published in The Lancet (2026). The Nature Reviews Clinical Oncology piece summarized these findings and emphasized that the phase III data confirm HCT–ASCT as a superior consolidation strategy in the selected, transplant‑eligible PCNSL population.
For full trial details, outcomes, safety profiles and subgroup results, refer to the original Lancet publication cited in the source highlight.