Effective transvascular delivery of nanoparticles across the blood-brain tumor barrier into malignant glioma cells

If you want to kill a brain tumor, you need a drug that can reach it. However, the brain is protected by one of the most selective barriers in biology. If you engineer a nanoparticle small enough to slip through that wall, you need to know exactly how small is small enough. The number that determines the entire problem is 11.7 nanometres. A team at the National Institutes of Health spent years running particles of precisely calibrated sizes through living tumors to find this measurement. Here's the core problem. Glioblastoma and other malignant gliomas are lethal in part because a physical barrier surrounds their blood vessels, blocking most drugs from entering. This barrier is called the blood-brain tumor barrier, or BBTB — a pathological variation of the normal blood-brain barrier that forms around tumor microvessels. Unlike the normal barrier, the BBTB has gaps: discontinuities in the endothelial cells lining the vessel walls. Those gaps are the only routes by which a systemically delivered drug can cross into tumor tissue — transvascular passage, from inside the blood vessel to the tumor space outside it. The critical question that had not been precisely answered was: how large are those gaps? The clinical consequences of not knowing are severe. Standard chemotherapies like temozolomide and carmustine are small enough to cross the BBTB, but they clear from the bloodstream so quickly that they never build up to lethal concentrations inside tumor cells. One workaround is to surgically implant carmustine-soaked polymer wafers directly into the resection cavity — effective locally, but only at the moment of surgery. Larger drug vehicles, like nanoparticles, stay in circulation longer, but they have mostly failed because the BBTB pores are narrow enough to block them. The field needed a specific number, a real measured physiologic pore size. That's exactly what Sarin and colleagues set out to find, and their experimental tool was elegant. Polyamidoamine, or PAMAM, dendrimers are a class of synthetic polymers built in precise generations. Each generation adds a predictable shell of branching molecules around a central core, and crucially, the diameter increases by only about one to two nanometres per generation. Generation one weighs about 1.4 kilodaltons, while generation eight weighs 233 kilodaltons. The number of surface amine groups doubles with each generation — eight at generation one, up to one thousand twenty-four at generation eight. Because the size increments are so small and uniform, you can treat generations one through eight as a calibrated ruler — a series of probes spaced roughly two nanometres apart, spanning a range from a few nanometres to over a dozen. To make them trackable, the team attached two kinds of labels. They conjugated gadolinium-DTPA — a standard MRI contrast agent — to the surface amines, which allowed them to watch the particles in real time using dynamic contrast-enhanced MRI, or DCE-MRI. That technique measures signal changes in tissue over time as the contrast agent arrives and accumulates — a direct readout of whether particles have crossed from the bloodstream into the tumor. They also labeled selected generations with rhodamine B, a fluorescent dye, so they could confirm particle location inside individual cells after the animals were sacrificed. The whole system was tested in rats with RG-2 gliomas grown directly in the brain, which represents an orthotopic model, meaning the tumors were in their natural location, with a functioning blood-brain tumor barrier. The DCE-MRI results were striking in their clarity. At the standard imaging dose of 0.03 millimoles of gadolinium per kilogram of body weight, generations one through five crossed the blood-brain tumor barrier and accumulated in the extravascular tumor space. Generations six, seven, and eight did not. At a higher dose — 0.09 millimoles per kilogram — generation six also crossed. Generation seven was variable: it remained intravascular in smaller tumors but could extravasate in the largest, most defective tumors. Generation eight was impermeable at both doses, in every tumor. Particle sizing by annular dark-field scanning transmission electron microscopy gave the physical correlate of that functional cutoff. Generation seven dendrimers averaged 11.0 nanometres in diameter, plus or minus 0.7. Generation eight averaged 13.3 nanometres, plus or minus 1.4. Sarin and colleagues placed the physiologic upper limit of blood-brain tumor barrier pore size at approximately 11.7 to 11.9 nanometres — marking the gap between the largest particle that could reliably cross and the smallest that reliably could not. Below that threshold, transvascular passage was possible. Above it, the barrier held. Pharmacokinetic modeling confirmed these results: across the permeable generations, the team found statistically significant differences in transvascular flow rate and in the fraction of extravascular space the particles occupied, while the fractional plasma volume did not differ — indicating the cutoff was related to the barrier, not to vascular architecture. But here's where the story gets more interesting. Getting through the barrier turned out to be only half the problem. Among the generations that could cross, not all of them actually accumulated inside glioma cells. The difference came down to how long the particles stayed in the blood. Smaller generations — from generation one through lowly conjugated generation four — were cleared quickly by the kidneys. They fall below the roughly 30 to 40 kilodalton renal filtration threshold, so they appear in the tumor space briefly and then disappear. Larger generations, from generation five and above, maintained steady blood concentrations for at least two hours after infusion. That sustained circulation meant sustained transvascular flux. Generation five showed a steady rate of accumulation in tumor tissue over two hours at both doses, accumulating faster in the first hour at the higher dose. Generation six only accumulated meaningfully at the higher dose. The pattern is consistent: a short blood half-life means the concentration gradient driving extravasation collapses before enough particles reach tumor cells. A long half-life keeps that gradient alive. The fluorescence data confirmed this directly. In living animals, rhodamine-labeled generation five dendrimers showed substantial accumulation and subcellular localization inside tumor cells when examined after sacrifice. Rhodamine-labeled generation eight — which couldn't cross the blood-brain tumor barrier — showed only minimal presence in tumor tissue and essentially no subcellular localization. In cell culture, generations two, five, and eight all entered RG-2 cells when added directly — indicating that the barrier to generation eight in vivo is the blood-brain tumor barrier itself, not the cell membrane. The blood-brain tumor barrier is the gate. The blood half-life is the key to using that gate effectively. So two rules emerge, and both are necessary. A nanoparticle designed to treat a brain tumor must be smaller than roughly 11.7 to 11.9 nanometres — that's the pore size ceiling. Additionally, it must be large enough, or engineered carefully enough, to avoid rapid renal clearance and maintain a long blood half-life. Threading that needle is the design challenge presented to the field by this research. The authors are careful to add one important caveat. Surface charge matters, and not just for performance. When rhodamine B was added as a label, it introduced cationic — positively charged — groups on the dendrimer surface, and this was associated with apparent toxicity and signs of barrier disruption. Cationic surface charge can damage cell membranes and disrupt the barrier itself. That means you can't simply maximize surface amines to force entry into tumors — doing so risks toxicity. The surface chemistry has to be tuned alongside the size and the half-life. What Sarin and colleagues have delivered is not just a measurement but a target. For decades, nanoparticle-based drug delivery to brain tumors was a problem without a quantitative boundary condition. Now there is one: 11.7 to 11.9 nanometres. That number applies to liposomes, quantum dots, iron oxide particles, antibody conjugates, and viral vectors — any platform the field wants to develop for transvascular delivery into malignant gliomas. Design the particle below the threshold, engineer the surface chemistry to achieve a long circulation time without toxicity, and you have a rational path to reaching tumor cells that conventional chemotherapy cannot. The brain's defenses are formidable, but they have a measurable limit. And now we know what it is. This lecture was created by ennepō. Go to https://ennepo.ai to Discover, Create and Follow the latest research in your field. Read when you can. Listen when you want to.