Can a Stem Cell Transplant Improve Survival in Primary CNS Lymphoma? What a New Phase 3 Trial Found

A large international Phase 3 trial known as MATRix/IELSG43, published in The Lancet in July 2026, found that consolidating primary central nervous system lymphoma (PCNSL) with high-dose chemotherapy and an autologous stem-cell transplant produced substantially better three-year survival and far fewer relapses than a non-transplant chemotherapy consolidation regimen. This article walks through what primary CNS lymphoma is, what the trial actually tested and found, the toxicity and eligibility trade-offs that come with transplant, and how this result fits into two decades of effort to find a safer, more durable way to treat this rare and historically difficult cancer.

What Is Primary CNS Lymphoma, and Why Has It Been So Hard to Treat?

Primary CNS lymphoma is a rare, aggressive form of non-Hodgkin lymphoma in which malignant B cells arise within the brain, spinal cord, cerebrospinal fluid, or eyes, without lymphoma elsewhere in the body. It's more common in people with weakened immune systems, though it also occurs in immunocompetent adults, and outcomes have historically been sobering: median survival in older trials ranged from roughly two to five years, and age over 60 has long been recognized as a poor prognostic factor.

Part of what makes PCNSL difficult is the blood-brain barrier, which limits which drugs can reach the tumor at effective doses. For years, the mainstay of treatment was high-dose methotrexate-based chemotherapy followed by whole-brain radiotherapy (WBRT). WBRT can control disease, but it comes with a well-documented cost: standard-dose WBRT (around 40 Gy) is associated with measurable declines in balance and cognitive function, especially in older patients, which pushed the field to look for consolidation strategies that could match its efficacy without the same neurological toll.

The MATRix/IELSG43 Trial: What Was Tested

MATRix/IELSG43 is described by its investigators as the largest randomized trial ever conducted in untreated PCNSL, run across 56 sites by the International Extranodal Lymphoma Study Group and the German Cooperative PCNSL Study Group. It enrolled 368 previously untreated, immunocompetent adults with B-cell PCNSL — patients up to 65 could enroll regardless of performance status, while those aged 66 to 70 needed a good baseline functional status (ECOG 0–2) to qualify.

All patients first received four cycles of MATRix induction chemotherapy (rituximab, high-dose methotrexate, cytarabine, and thiotepa). Of the 346 who started induction, 230 responded well enough and were fit enough to be randomized (229 included in the final analysis) to one of two consolidation strategies: two cycles of R-DeVIC, a non-myeloablative rituximab-based chemotherapy regimen, or high-dose chemotherapy followed by an autologous stem-cell transplant.

The Results: A Clear, Well-Powered Win for Transplant Consolidation

The results were both statistically strong and clinically meaningful. At a median follow-up of 45.3 months, patients randomized to transplant consolidation had a three-year progression-free survival of 78%, compared with 51% for R-DeVIC (hazard ratio 0.43, p=.0003) — median PFS wasn't even reached in the transplant group, versus about 39 months with chemotherapy alone. Three-year overall survival followed the same pattern: 86% with transplant versus 71% with R-DeVIC (hazard ratio 0.46, p=.0075). Perhaps most striking for patients and families weighing this decision, the three-year relapse rate was cut more than in half, from 45% with R-DeVIC down to 19% with transplant. Notably, the two groups had similar early response to treatment (complete response rates of 69% versus 65% at day 60), which underscores that the real advantage of transplant consolidation lies in how durable that response turns out to be, not in getting patients into remission in the first place.

The Trade-offs: Toxicity, Treatment-Related Deaths, and Who Actually Qualifies

None of this comes without cost, and the trial's own investigators were explicit that "this highly intensive treatment approach is not universally applicable." Patients who received transplant consolidation experienced more adverse events overall (an average of 14.6 per patient versus 9.3), with febrile neutropenia, stomatitis, and diarrhea occurring more often during the transplant process. Treatment-related deaths were also more frequent in the transplant arm: five deaths, four attributed to infectious complications and one to a pulmonary embolism (the embolism was not judged treatment-related, but the others were). In the chemotherapy-only arm, there were two treatment-related deaths, both from secondary acute myeloid leukemia, a recognized long-term risk of the alkylating and topoisomerase-inhibiting drugs used in this disease.

Just as important is who never made it into the randomized comparison at all: of the 346 patients who started induction, 116 (34%) did not proceed to randomization, largely because of treatment-related toxicity or because their disease progressed or failed to respond well enough during induction. In other words, these encouraging numbers describe outcomes for the roughly two-thirds of patients who were fit enough and responded well enough to reach the point of choosing a consolidation strategy — not for every person diagnosed with PCNSL.

How This Fits What We Already Knew

This trial builds directly on an earlier study, IELSG32, which in 2017 first showed that autologous stem-cell transplant could match whole-brain radiotherapy's efficacy as a consolidation option after methotrexate-based induction — and did so with a notable advantage: cognitive testing showed WBRT patients tended to decline in attention and executive function, while ASCT patients in that trial showed improvements in memory, cognitive function, and quality of life. Seven-year follow-up of that same trial reported roughly 70% overall survival with MATRix induction plus consolidation, and no excess of non-relapse mortality or second cancers attributable to the transplant itself. MATRix/IELSG43 takes the next logical step, testing transplant not against radiotherapy but against a non-myeloablative chemotherapy regimen, and it does so as a fully powered Phase 3 trial rather than a randomized comparison nested in a Phase 2 study. That progression — from "transplant is at least as good as radiation, with a better cognitive profile" to "transplant beats non-transplant chemotherapy consolidation on hard survival endpoints" — is what makes this result a meaningful advance rather than an incremental one.

What This Means for Patients Today

For eligible patients — generally those healthy enough to tolerate intensive induction chemotherapy and to reach the consolidation phase in a good response — this trial gives oncologists solid, randomized evidence to support autologous stem-cell transplant as the preferred consolidation strategy, a shift already reflected in commentary accompanying the publication. It does not change the calculus for patients who can't tolerate intensive induction, who are older with reduced functional status, or whose disease doesn't respond well enough to reach randomization in the first place; for them, other approaches, including lower-intensity chemotherapy, reduced-dose radiotherapy, or emerging targeted and immunotherapy options, remain the relevant conversation.

Bottom Line

If you or a loved one has been newly diagnosed with primary CNS lymphoma, this trial is a genuine reason for optimism, but it applies specifically to patients who are fit enough to complete intensive induction chemotherapy and respond well enough to be considered for transplant. Ask your oncology team directly whether you meet that profile, what your individual risk of transplant-related complications (including infection and, rarely, fatal ones) looks like given your age and health, how your case compares to the trial's enrollment criteria, and what alternative consolidation options exist if transplant isn't appropriate for you. This is a treatment-planning decision that should be made with a lymphoma specialist familiar with the newest data, not from this article alone.