Does Chemotherapy Cross the Blood-Brain Barrier?

Most chemotherapy drugs cross the blood-brain barrier poorly or not at all, which is a central reason why brain tumors remain among the hardest cancers to treat. The barrier is not a single wall but a set of overlapping defenses, from physical seals between cells to protein pumps that actively eject foreign molecules back into the bloodstream. A handful of drugs do get through in meaningful amounts, and oncologists have developed several creative workarounds for the rest. But the basic challenge persists: the very system that protects your brain from toxins also shields tumors growing inside it.

What Makes the Blood-Brain Barrier So Effective

The blood vessels that supply your brain are fundamentally different from blood vessels elsewhere in the body. In most organs, the cells lining capillaries have small gaps and fenestrations that let molecules slip through. Brain capillaries do not. Their endothelial cells are locked together by continuous tight junctions that create an exceptionally high electrical resistance, making it very difficult for water-soluble molecules to squeeze between cells.¹

But tight junctions are only part of the story. Even if a drug molecule manages to enter a brain endothelial cell, it often gets pumped right back out. The barrier is equipped with efflux transporters, proteins embedded in the cell membrane that recognize a wide range of foreign molecules and actively shuttle them out of the brain. The most studied of these is P-glycoprotein, sometimes called P-gp, which handles many common chemotherapy drugs.² Other efflux transporters, including multidrug-resistance-associated proteins, work alongside P-gp to create a multi-layered chemical defense system.¹

A drug’s ability to cross also depends on its chemical properties. Compounds that are moderately fat-soluble tend to have the best shot at getting in, because they can dissolve into the lipid-rich cell membranes of the barrier. But the relationship is not straightforward: very fat-soluble drugs tend to stick to blood proteins and get cleared faster by the liver, while very water-soluble ones get filtered out by the kidneys before they reach the brain in significant concentrations.³ Size matters too: smaller molecules slip through more easily than large ones. The result is a narrow sweet spot that relatively few chemotherapy agents happen to occupy.

Which Drugs Get Through and Which Do Not

Temozolomide, the standard chemotherapy for glioblastoma, is one of the few that does cross the barrier in clinically relevant amounts. It is small, moderately lipophilic, and can reach tumor tissue in the brain after being swallowed as a pill. Even so, efflux pumps still limit how much arrives. In mice lacking both the P-gp and BCRP efflux transporters, brain concentrations of temozolomide were about 50% higher than in normal mice, while blood levels stayed the same.⁴ That finding underscores an uncomfortable truth: even the drugs we consider “brain-penetrant” are still being partially ejected by the barrier.

High-dose methotrexate and high-dose cytarabine also reach the central nervous system when given at doses far above what would be used for cancers elsewhere in the body. Both are standard tools for treating leukemias that threaten to invade the brain and for primary central nervous system lymphoma.⁵ The strategy of simply flooding the bloodstream with enough drug to overwhelm the barrier works, but it comes with significant side effects in the rest of the body.

Many widely used agents, however, are essentially locked out. Doxorubicin, one of the most effective drugs against a broad range of cancers, barely penetrates the brain because efflux transporters recognize and remove it aggressively.⁶ First-generation taxanes, including paclitaxel and docetaxel, are similarly unable to cross.⁷ Platinum agents like cisplatin and carboplatin have limited brain penetration on their own as well, which is why they are often combined with barrier-disruption techniques when used for brain tumors.

The Blood-Tumor Barrier Is Not the Same as the Blood-Brain Barrier

One common assumption is that brain tumors simply sit behind an intact barrier, but the reality is messier. As tumors grow, they recruit new blood vessels through a chaotic process that often produces vessels structurally different from normal brain capillaries. The resulting blood-tumor barrier can be partially “leaky,” allowing some drug molecules in that would otherwise be excluded. This is why contrast dye lights up many brain tumors on an MRI scan: the dye is leaking through abnormal vessels that a healthy barrier would block.

The trouble is that leakiness is uneven. A 2024 single-cell study of human glioma tissue found that in lower-grade tumors, the blood vessels looked essentially identical to normal brain tissue at the molecular level, while glioblastoma vessels showed a range of abnormalities.⁸ So one region of a high-grade tumor might let chemotherapy seep in while a neighboring region keeps it out. This heterogeneity is a major obstacle: even when a drug technically “crosses” into some parts of a brain tumor, it may not reach all of the tumor cells, especially the invasive ones at the margins that tend to cause recurrences.⁹ And even in leaky regions, efflux pumps can still be active, pumping drug back out as fast as it leaks in.

Ways Oncologists Get Around the Barrier

Because so many effective drugs are excluded, a significant part of neuro-oncology research focuses on getting around or through the barrier rather than finding new drugs. Several approaches are in clinical use or active trials.

Intrathecal Chemotherapy

The most direct bypass is to skip the bloodstream entirely and inject drugs straight into the cerebrospinal fluid. Intrathecal chemotherapy, usually methotrexate or cytarabine delivered via lumbar puncture or an implanted reservoir, is routine for preventing or treating leukemia and lymphoma in the central nervous system.⁵ In acute lymphoblastic leukemia, virtually all patients receive intrathecal prophylaxis because the risk of the cancer hiding in the brain is high. For acute myeloid leukemia, the decision is more individualized: a large retrospective study of 960 patients found that prophylactic intrathecal methotrexate was associated with a lower five-year rate of brain relapse compared to no prophylaxis, though the benefit may be most relevant for patients at higher risk.¹⁰

Implanted Wafers

For solid brain tumors, surgeons can place drug-releasing implants directly into the cavity left after a tumor is removed. The only FDA-approved version is the carmustine wafer, a biodegradable polymer disc that slowly releases a chemotherapy agent into the surrounding brain tissue over weeks.¹¹ The concept is appealing: the drug never has to cross the barrier because it is already on the brain side. In practice, the wafers have modest survival benefits and limited drug penetration beyond a few millimeters from the implant surface, so research is ongoing into next-generation local delivery systems.

Osmotic Disruption

A more aggressive approach is to temporarily force the barrier open. Injecting a concentrated sugar solution, typically mannitol, into a brain artery causes the endothelial cells to shrink, widening the tight junctions long enough for chemotherapy to flood through.¹² This technique has been used to deliver methotrexate, rituximab, and carboplatin for primary central nervous system lymphoma.¹³ It works, but it requires a specialized interventional radiology team and general anesthesia, and the barrier opens indiscriminately, not just over the tumor, so healthy brain tissue gets exposed to chemotherapy as well.

Focused Ultrasound

A newer and more targeted method uses focused ultrasound combined with injected microbubbles to open the barrier in a precise, image-guided region. The ultrasound waves cause the microbubbles to oscillate against vessel walls, temporarily loosening tight junctions in the targeted area. MRI guidance lets clinicians see exactly where the barrier is being opened and confirm that it closes again afterward.¹⁴ Early clinical trials are underway for glioblastoma and brain metastases, and PET imaging is being used to track how much drug actually reaches the brain after ultrasound-mediated opening.¹⁵

Nanoparticles

Researchers are also engineering tiny drug-carrying particles designed to exploit the barrier’s own transport systems. Nanoparticles can be coated with molecules that mimic substances the brain normally imports, essentially tricking the barrier’s receptor-mediated transport machinery into pulling the drug-loaded particle across.¹⁶ Most nanoparticle platforms are still in preclinical or early clinical development, but they represent one of the most active areas of brain tumor research.

Does Radiation Open the Barrier?

Patients with brain tumors almost always receive radiation, and a reasonable question is whether radiation itself makes the barrier more permeable to chemotherapy given afterward. The answer is technically yes, but probably not enough to matter. A meta-analysis of 29 preclinical studies confirmed that radiation does increase barrier permeability in animal models.¹⁷ However, when researchers measured actual drug distribution in mouse brains after therapeutic doses of radiation, the effects on drug delivery were small, short-lived, and unlikely to be clinically meaningful.¹⁸ In other words, radiation nudges the door open slightly, but not enough to reliably improve chemotherapy delivery. This is worth knowing because patients sometimes assume that combining radiation with chemo works partly by letting more drug into the brain; the benefit of the combination is more likely due to each treatment attacking tumor cells through different mechanisms.

Chemo Brain and What It Tells Us About Barrier Permeability

If most chemotherapy cannot cross the blood-brain barrier, why do so many cancer patients experience cognitive problems during and after treatment? The phenomenon, widely known as “chemo brain,” involves memory lapses, difficulty concentrating, and mental fogginess that can persist for months or years. Reported rates vary widely, affecting anywhere from about one in six to three in four patients depending on the drug regimen and how cognitive impairment is measured.¹⁹

It might seem contradictory: drugs that cannot cross the barrier still affect the brain. But current research suggests the mechanism is indirect. Chemotherapy appears to trigger persistent activation of microglia, the brain’s resident immune cells, even when the drug itself stays in the bloodstream.²⁰ Inflammatory signals from the body can cross the barrier even when the drug molecules cannot, and once microglia become activated, they release their own inflammatory molecules inside the brain that disrupt the normal processes underlying memory and learning. In mouse studies, paclitaxel, a drug that does not cross the barrier well, still triggered a measurable inflammatory response in the hippocampus, the brain region critical for forming new memories. When researchers depleted the mice’s microglia, the cognitive impairment went away, strongly suggesting that the immune cells rather than the drug itself were driving the problem.²¹

Separate research has also proposed that chemotherapy-induced oxidative damage and epigenetic changes in the brain may overlap with aging processes, which could help explain why older patients sometimes experience more pronounced cognitive effects.¹⁹ The take-home point is that chemo brain is not evidence that drugs are flooding the brain; it is evidence that the body’s response to chemotherapy can affect the brain through immune and inflammatory signaling pathways that operate independently of the barrier.

Genetic Variation in Efflux Pumps

Not everyone’s blood-brain barrier pumps drugs out at the same rate. The gene ABCB1 encodes P-glycoprotein, and common variations in this gene have been studied as potential explanations for why some brain tumor patients respond better to chemotherapy than others. One early study of glioblastoma patients treated with temozolomide found that people carrying a specific variant had a two-year survival rate of 37%, compared to roughly 8-10% in those with other genotypes.²² That is a striking difference, and it generated considerable excitement.

However, follow-up studies failed to replicate a consistent survival benefit linked to ABCB1 genotype. A recent systematic review concluded that while some variants do appear to affect P-glycoprotein function, any effect on temozolomide response is overshadowed by more powerful factors like the tumor’s own molecular profile.²³ Mouse studies confirm that certain ABCB1 mutations can reduce P-glycoprotein’s pumping efficiency and increase brain penetration of its substrates, even without changing how much of the protein is produced.²⁴ So the biology is real, but translating it into treatment decisions for individual patients has not proven reliable.

How the Barrier Differs in Children

Brain tumors are the most common solid tumors in children, and the barrier adds a layer of complexity that is distinct from what adult patients face. A 2025 review found that barrier integrity varies significantly across pediatric brain tumor types and is shaped by the tumor’s molecular subtype and its location in the brain.²⁵ Low-grade gliomas in children tend to have a relatively intact barrier, making drug delivery especially difficult. Medulloblastomas and certain high-grade tumors may have regions of increased permeability, but as in adults, this leakiness is uneven and unpredictable.

The developing brain also raises safety concerns that do not apply to the same degree in adults. A child’s brain is still forming connections, pruning synapses, and myelinating nerve fibers. Strategies that aggressively disrupt the barrier, like osmotic opening, carry risks of exposing developing neural circuits to toxic compounds. This is one reason why first-generation taxanes, which cannot cross the barrier, have been largely ineffective in pediatric brain tumor patients, and why newer agents like cabazitaxel, designed to evade efflux pumps, are being studied specifically for this population.⁷ Tailoring treatment to the specific barrier characteristics of each tumor type, rather than assuming a one-size-fits-all level of permeability, is increasingly seen as essential in pediatric neuro-oncology.

Measuring What Actually Gets Through

For decades, most of what we knew about drug penetration into the brain came from animal studies or from measuring drug concentrations in cerebrospinal fluid, which is an imperfect proxy for what is actually happening inside brain tissue. Newer imaging techniques are changing this. A novel PET-based method now allows researchers to measure the permeability of the barrier to specific molecules in living humans during a single scan, capturing not just whether the barrier is structurally intact but how it handles particular molecular tracers passing through different transport routes.²⁶ This kind of molecular-level measurement could eventually let clinicians predict, before starting treatment, whether a specific drug will reach a specific patient’s tumor in adequate concentrations. The gap between knowing a drug crosses the barrier in a mouse model and knowing it reaches the tumor in a particular patient’s brain is large, and closing it with better imaging stands to make brain tumor treatment considerably more precise.