High-dose methotrexate failed to reduce CNS relapse in ultra-high-risk large B-cell lymphoma, raising fresh doubts about one of the field's most widely used preventive strategies.
High-dose methotrexate does not prevent central nervous system relapse in the highest-risk patients with large B-cell lymphoma, challenging one of the most widely used but controversial treatment strategies, an international study co-led by Australian researchers has found.
For years, clinicians have accepted the toxic trade-off of high-dose methotrexate in the hope of preventing devastating central nervous system (CNS) relapse in patients with ultra-high-risk large B-cell lymphoma.
Published in the Journal of Clinical Oncology, the multicentre retrospective analysis found no evidence that adding high-dose methotrexate as CNS prophylaxis reduced the risk of CNS relapse or improved survival in patients considered at the very highest risk of disease spreading to the brain or spinal cord.
“To our knowledge, this is the largest international comparative data set of patients with LBCL with UHR features showing no significant reduction in CNS relapse with HD-MTX in all UHR or in any individual subgroup,” the researchers wrote.
“Despite inherent limitations of retrospective analyses, the data strongly support previous studies in suggesting HD-MTX prophylaxis has no meaningful benefit for most patients.”
The findings are expected to reignite debate over the role of prophylactic high-dose methotrexate, a practice that has remained common despite a lack of prospective randomised evidence and growing concerns about toxicity, treatment delays, and uncertain benefit.
The study was co-led by Dr Katharine Lewis from Blood Cancer Research WA and involved investigators from Australia, Europe, and North America.
Dr Lewis, a haematologist and clinician-researcher, said the findings helped answer an important clinical question that has been debated for many years.
“For patients with large B-cell lymphoma, the possibility of the disease spreading to the brain or spinal cord is one of the greatest concerns,” she said.
“Our study provides important evidence to help clinicians and patients have informed conversations about the potential benefits and limitations of current preventative treatment, while also highlighting the need for better strategies to reduce the risk of CNS relapse.”
The researchers analysed outcomes for patients with large B-cell lymphoma classified as ultra-high risk for CNS relapse, comparing those who received high-dose methotrexate prophylaxis with those who did not.
Despite targeting the group thought most likely to benefit, the investigators found no significant reduction in CNS relapse among patients who received prophylactic methotrexate. Overall survival and progression free survival were also not improved.
The results challenge a longstanding assumption in lymphoma management. CNS relapse occurs in only a small proportion of patients with diffuse large B-cell lymphoma but is associated with an extremely poor prognosis, with survival often measured in months once relapse occurs.
Because of these devastating outcomes, clinicians have sought ways to prevent CNS involvement in patients deemed to be at elevated risk based on clinical features, extranodal disease sites, and CNS International Prognostic Index scores, the researchers wrote.
High-dose intravenous methotrexate has become one of the most widely adopted approaches despite inconsistent evidence supporting its effectiveness.
Related
The treatment is not without cost. High-dose methotrexate requires inpatient administration, intensive hydration and monitoring, and carries risks including renal toxicity, mucositis, hepatotoxicity, and treatment delays that can interfere with curative systemic chemotherapy.
These concerns have fuelled increasing scepticism in recent years, with several observational studies questioning whether the treatment meaningfully alters the risk of CNS relapse. However, evidence has remained limited by heterogeneous patient populations and varying definitions of risk.
The current study sought to address this uncertainty by focusing specifically on patients considered to be at ultra-high risk, representing the group most likely to derive benefit if prophylaxis were effective. Instead, researchers found no clinically meaningful advantage associated with methotrexate prophylaxis across the measured outcomes.
Blood Cancer Research WA described the work as a landmark international collaboration that could reshape clinical practice by prompting clinicians to reconsider the routine use of high-dose methotrexate for CNS prevention in this setting.
Professor Chan Cheah, a specialist physician in haematology and pathology, clinical researcher, founder of Blood Cancer Research WA, and a study investigator said studies that challenged long-held assumptions were often among the most important.
“The role of CNS prophylaxis in high-risk large B-cell lymphoma has been debated for many years,” he said.
“This work provides valuable evidence to help guide discussions with patients and inform clinical practice.”
The study is unlikely to end the debate immediately. As a retrospective analysis, it cannot completely eliminate the influence of treatment selection bias, and prospective randomised trials would provide stronger evidence.
However, conducting such trials has proven difficult because CNS relapse is relatively uncommon, even among high-risk patients, meaning very large studies would be required.
For now, the data add to a growing body of evidence suggesting that clinicians should carefully weigh the substantial toxicity and logistical burden of high-dose methotrexate against increasingly uncertain evidence of benefit, the researchers wrote.
In an accompanying editorial, the Journal of Clinical Oncology’s editor in chief Dr Jonathan W. Friedberg, said that despite these limitations, when taken in context of other studies, he felt “these results should definitively end the use of high-dose methotrexate as CNS prophylaxis for DLBCL”.
“It should be emphasised that methotrexate remains an important component of treatment for high-grade lymphomas, especially Burkitt lymphoma,” Dr Friedberg wrote.
“In DLBCL, a definitive prospective randomized trial of methotrexate would be very challenging to conduct given the low event rate even in the highest risk patients. Given the preponderance of evidence demonstrating limited or no activity of methotrexate in preventing CNS relapse, such a trial should not be pursued.
“Future efforts need to focus on improving risk stratification to better define the group of patients with the highest risk of CNS progression and to study novel agents with early signals of efficacy in CNS lymphoma.
“The ability to measure circulating tumour DNA in CSF represents an opportunity to define patients with occult CNS involvement of DLBCL, who comprise a significant subset of the patients who ultimately relapse with CNS disease.”
He said the authors should be congratulated on their “robust analysis, which demonstrates the power of international collaborations within our lymphoma community.
“As our most common lymphoma (DLBCL) is now being divided into molecular subsets within a rapidly changing treatment landscape, such international collaborations become even more essential,” Dr Friedberg wrote.
Attention was also shifting towards better methods of identifying which patients are genuinely at risk of CNS relapse. Advances in molecular profiling, circulating tumour DNA, imaging, and biomarker research may eventually enable more precise risk stratification than current clinical scoring systems the study researchers wrote.
Researchers are also investigating whether emerging targeted therapies, bispecific antibodies, and cellular immunotherapies could reduce CNS relapse risk through improved systemic disease control rather than relying on separate prophylactic strategies.
“In conclusion, we report what is to our knowledge the largest international data set of patients with LBCL with UHR features, demonstrating no significant reduction in CNS relapse with HD-MTX prophylaxis,” the researchers wrote.
“There is an urgent need to design prospective, randomised trials aimed exclusively at UHR patients, using novel therapies other than HD-MTX.”
Journal of Clinical Oncology, June 2026 (paper)
Journal of Clinical Oncology, June 2026 (editorial)
