Primary CNS Lymphoma Treatment: Methotrexate and CAR T

Treatment for primary central nervous system lymphoma (PCNSL) revolves around high-dose methotrexate-based chemotherapy as the backbone of initial therapy, followed by a consolidation phase that increasingly favors autologous stem-cell transplant over whole-brain radiation. This two-phase approach has meaningfully improved outcomes over the past two decades, but managing PCNSL remains more complicated than treating lymphomas elsewhere in the body, largely because most drugs struggle to cross the blood-brain barrier. The treatment decisions start even before chemotherapy begins, and the right choices at each stage can significantly affect both survival and long-term quality of life.

Why Getting the Diagnosis Right Matters So Much

Before any treatment can start, PCNSL has to be properly diagnosed, and a few neurosurgical principles make this trickier than it sounds. The standard approach is a stereotactic biopsy rather than attempting to remove the tumor entirely. Unlike many other brain tumors, PCNSL does not benefit from aggressive surgical resection, so taking out as much tumor as possible offers no survival advantage and only adds surgical risk.¹

A major pitfall involves corticosteroids. Steroids are commonly given to patients with brain lesions to reduce swelling, but in PCNSL they can temporarily shrink or even seem to dissolve the tumor, making it difficult to get enough tissue for a diagnosis. This “vanishing tumor” effect can lead to inconclusive biopsies, delaying treatment. If PCNSL is suspected, steroids should be held until after the biopsy whenever safely possible.² That said, if a patient is already on steroids and the tumor is still visible and large enough to biopsy, the procedure can still succeed. Doctors weigh the tumor’s size, location, and how much it has responded to steroids before deciding whether to proceed.³

High-Dose Methotrexate as the Foundation

Once PCNSL is confirmed, the mainstay of initial treatment is high-dose methotrexate (HD-MTX). Methotrexate is one of the few chemotherapy drugs that can penetrate the blood-brain barrier when given at high enough doses, which is why it occupies a central role that no other agent has displaced. A large meta-analysis pooling data from 37 clinical studies found that overall response rates were roughly 78–80% across different methotrexate dose levels, but survival separated more clearly: two-year overall survival was about 52% at lower doses, 60% at medium doses, and 71% at higher doses.⁴ This dose-response relationship is one reason most centers aim for at least 3 grams per square meter of body surface area per cycle.

Methotrexate is rarely given alone. The most studied multi-drug regimen is called MATRix, which adds cytarabine, thiotepa, and the antibody drug rituximab to methotrexate. A landmark randomized trial compared three induction arms head-to-head. Patients who received the full MATRix combination had a seven-year overall survival of 56%, compared with 37% for those who received methotrexate, cytarabine, and rituximab without thiotepa, and just 21% for methotrexate and cytarabine alone. When patients who received MATRix went on to consolidation therapy, seven-year survival rose to 70%.⁵ The trial also showed that adding rituximab to the backbone of methotrexate and cytarabine already improved outcomes, supporting the idea that rituximab plays a meaningful role even inside the brain.

Consolidation After Induction

Getting a good initial response is only half the battle. Without some form of consolidation therapy, many patients relapse within months. The two major consolidation strategies are whole-brain radiotherapy (WBRT) and high-dose chemotherapy followed by autologous stem-cell transplant (ASCT). The choice between them has become one of the more consequential decisions in PCNSL care.

For years, WBRT was the default. It is effective at killing residual tumor cells throughout the brain, but it comes at a steep long-term cost. When given at standard doses above 30 Gy, WBRT carries a real risk of permanent, irreversible damage to thinking and memory, especially in older adults.⁶ A long-term follow-up of the PRECIS trial, which randomized patients aged 60 and younger to either ASCT or 40 Gy WBRT, found that eight-year event-free survival was 67% with ASCT versus 39% with WBRT. The risk of relapse was dramatically lower after transplant. Balance problems developed in more than half of WBRT-treated patients compared with about 10% of transplant patients, and thinking skills deteriorated in roughly two-thirds of the WBRT group versus about 13% of the transplant group.⁷

A more recent phase 3 trial further solidified the case for transplant. In the MATRix/IELSG43 trial, patients who responded to MATRix induction were randomized to either high-dose chemotherapy with ASCT or a non-myeloablative regimen. After a median follow-up of about 45 months, three-year progression-free survival was 78% with transplant compared with 51% without it.⁸ For fit patients who respond well to induction, ASCT has become the preferred consolidation approach at most major centers.

A retrospective analysis from China echoed these findings, reporting two-year progression-free survival of about 89% and four-year overall survival of roughly 71% in patients who received transplant consolidation.⁹ That study also identified early achievement of complete or partial remission as an independent predictor of better outcomes, reinforcing the importance of a strong response during induction.

Reduced-dose WBRT and other non-transplant consolidation regimens still have a role, particularly for patients who are not candidates for the intensive conditioning that transplant requires. Some protocols use high-dose cytarabine and etoposide as consolidation, which was shown to be well tolerated with no treatment-related deaths or signs of brain toxicity in an early pilot study.¹⁰ There is also ongoing work to identify lower radiation doses that preserve some of WBRT’s antitumor activity while minimizing brain injury.

When the Lymphoma Comes Back

Despite improvements in first-line treatment, a significant proportion of patients either relapse or never fully respond. Relapsed or refractory PCNSL is difficult to treat, but a new class of drugs called BTK inhibitors has opened a meaningful avenue. These drugs target Bruton’s tyrosine kinase, an enzyme that is important in the signaling pathways that keep PCNSL cells alive. A meta-analysis of BTK inhibitors in relapsed or refractory PCNSL found a pooled overall response rate of about 70–72%, with complete remission in roughly 43–50% of patients.¹¹¹²

The catch is that BTK inhibitor monotherapy, while it produces responses, tends to yield short-lived remissions. One review of ibrutinib in relapsed PCNSL reported overall response rates ranging from about 52% to 89%, but median progression-free survival was only around 4.6 to 4.8 months.¹³ Combining BTK inhibitors with chemotherapy or radiation appears to improve results substantially. In one meta-analysis, the complete remission rate jumped from 7% with monotherapy to 68% when combined with chemotherapy and 80% with radiation.¹⁴ This strongly suggests that BTK inhibitors work best as part of a multi-agent approach rather than as standalone treatment.

The molecular profile of the tumor may also predict who benefits most. A subtype of PCNSL characterized by mutations in two genes, MYD88 and CD79B (known as the MCD subtype), appears particularly sensitive to these targeted agents. A retrospective study of zanubrutinib combined with lenalidomide and rituximab found that patients with the MCD subtype had significantly better progression-free survival than those without these mutations.¹⁵ Early-phase studies support the idea that the B-cell receptor signaling pathway involving these genes is central to how PCNSL survives, though resistance mechanisms still need to be better understood.¹⁶

CAR T-Cell Therapy for PCNSL

Chimeric antigen receptor (CAR) T-cell therapy, which re-engineers a patient’s own immune cells to hunt cancer, has generated considerable excitement in PCNSL. There was initially concern that it might cause dangerous brain swelling in a disease already situated in the central nervous system, but real-world data have been cautiously encouraging. A meta-analysis of 128 patients found that about 64% of those with PCNSL responded to CAR T-cell therapy, with 56% achieving a complete response. Cytokine release syndrome (a systemic inflammatory reaction) occurred in about 70% of patients, though severe cases were limited to roughly 13%. Immune-related brain toxicity of any grade occurred in about 53%.¹⁷

The largest reported cohort of PCNSL patients treated with CAR T-cells comes from the French LOC network, which treated 25 heavily pretreated patients (median of three prior lines of therapy). Complete response was achieved in 64% of patients, and one-year progression-free survival was 43%, with a plateau afterwards suggesting some durable remissions. Among those who had at least a partial response at the time of CAR T-cell infusion, one-year relapse-free survival reached 79%.¹⁸ For patients who have run out of conventional options, these results represent a genuine step forward, though wider access and longer follow-up are still needed.

Treating Older Adults

Age is one of the sharpest dividing lines in PCNSL treatment. The median age at diagnosis is in the mid-60s, yet most of the landmark trials that established current treatment standards enrolled younger, fitter patients. Older adults face a double problem: they tolerate intensive chemotherapy and transplant conditioning less well, and they are more vulnerable to the neurotoxicity of whole-brain radiation.

An individual-patient-data meta-analysis of elderly patients found that about three-quarters received HD-MTX-based therapy, and that this approach was associated with improved survival compared with other strategies. Performance status was the strongest predictor of outcomes, with patients who were reasonably functional having about half the mortality risk of those who were debilitated. Interestingly, more aggressive multi-drug HD-MTX regimens did not clearly outperform simpler HD-MTX with oral chemotherapy in older populations.¹⁹ This suggests that for older adults, a gentler but well-delivered methotrexate-based approach can be the sweet spot.

A prospective study that stratified patients by age found that outcomes dropped sharply for those over 70. Five-year overall survival was 42% for patients under 61, 31% for those aged 61–70, and only 17% for those over 70, despite dose adjustments.²⁰ Some recent approaches have shown promise in treating older patients with sequential methotrexate-based regimens without either WBRT or transplant, aiming to preserve brain function while still controlling the disease.²¹

When the Eyes Are Involved

PCNSL can involve the eyes, a presentation sometimes called primary vitreoretinal lymphoma. The eyes are technically part of the central nervous system, so lymphoma cells can settle in the vitreous humor or retina either as a primary site or alongside brain disease. Treatment typically combines systemic high-dose methotrexate with direct injection of methotrexate into the eye (intravitreal therapy) to address bilateral involvement, though whether this combination delays brain relapse remains debated.²² For younger patients with relapsed ocular disease, intensive consolidation followed by transplant is an option, and newer targeted drugs like ibrutinib and lenalidomide have shown activity in this setting as well.

Getting Drugs Past the Blood-Brain Barrier

One of the core challenges in PCNSL treatment is that the blood-brain barrier blocks most chemotherapy drugs from reaching the tumor in effective concentrations. A technique called blood-brain barrier disruption (BBBD) uses an injection of concentrated mannitol into the arteries feeding the brain, which temporarily opens tight junctions between cells lining blood vessels and allows chemotherapy to flood in at much higher concentrations than would otherwise be possible.

This approach has been studied over several decades. An early series of 30 patients treated with BBBD and intra-arterial methotrexate showed median survival of nearly 45 months in the more recent cohort, with preservation of thinking skills in patients who achieved complete responses without radiation.²³ A larger multi-center series confirmed durable tumor control with this strategy.²⁴ More recently, a phase II study combined BBBD with high-dose chemotherapy and transplant, achieving complete responses in 88% of patients, with two-year overall survival of about 67%.²⁵

The procedure is technically demanding: it requires neuro-interventional expertise, general anesthesia, and careful monitoring. It is available at only a handful of specialized centers. Still, for patients at those centers, BBBD offers an alternative way to intensify therapy without relying on radiation.

Managing Methotrexate Toxicity

High-dose methotrexate, while essential, requires hospital admission for aggressive hydration and monitoring of drug clearance through blood tests every six to twelve hours. When methotrexate lingers in the body longer than expected, the risk of kidney damage, mouth sores, and dangerously low blood counts rises. An enzyme called glucarpidase can rapidly break down methotrexate in the bloodstream. It is typically reserved as a rescue agent for patients whose kidneys fail to clear the drug, but a phase I study explored giving low fixed doses of glucarpidase routinely 24 hours after each methotrexate infusion to speed clearance. The 2,000-unit dose reduced plasma methotrexate by over 95% within 15 minutes in nearly all treatments administered. The study found no evidence that this rapid systemic clearance interfered with the drug’s effectiveness inside the brain, since methotrexate appears to be sequestered in the central nervous system compartment independently of blood levels.²⁶ If validated in larger trials, routine glucarpidase could shorten hospital stays and reduce the systemic side effects of methotrexate without sacrificing its anticancer effect in the brain.

Long-Term Brain Health and Quality of Life

For the growing number of PCNSL survivors, what happens to brain function and overall quality of life after treatment is not an afterthought. A study comparing long-term survivors found that those who had received WBRT scored significantly worse on tests of attention, executive function, and motor skills compared with survivors treated without radiation. Brain MRI scans revealed that white-matter abnormalities after WBRT were more than double those seen in any non-radiation group, and these imaging changes tracked with worse thinking skills and quality-of-life scores.²⁷

Outcomes look different when radiation is avoided. A five-year follow-up from the HOVON 105 trial, in which patients received chemotherapy without radiation, found that thinking and memory generally improved after treatment and remained stable for up to five years. About half of patients showed meaningful memory improvement at five years compared with their pre-treatment baseline. The notable exception was motor speed, which deteriorated in nearly half of patients by the four- to five-year mark. Fatigue also remained a persistent problem: while it improved after treatment, fatigue scores never returned to what would be considered normal.²⁸ Global quality-of-life scores and social functioning remained stable in the majority, with roughly 70–90% of patients maintaining or improving these measures at each assessment point.

These findings underscore why the field has moved so decisively away from standard-dose WBRT in first-line treatment. For younger patients, transplant-based consolidation preserves brain function far better. For older patients who cannot undergo transplant, chemotherapy-only strategies or reduced-dose radiation may offer a more tolerable path, even if they come with a modestly higher relapse risk.

HIV-Associated PCNSL

PCNSL was once closely associated with advanced HIV infection, and before effective antiretroviral therapy it was almost uniformly fatal. The landscape has changed. People living with well-controlled HIV can now be treated with the same methotrexate-based regimens used in the general population, including transplant consolidation. Case reports have documented successful treatment with rituximab and methotrexate-based chemotherapy followed by autologous stem-cell transplant in patients with AIDS-associated PCNSL.²⁹ The key prerequisite is adequate immune recovery on antiretroviral therapy; without it, the risk of treatment-related infection is prohibitive. As HIV management continues to improve, the gap between outcomes for HIV-positive and HIV-negative patients with PCNSL has been narrowing.