Korean Scientists Create Chip That Predicts Brain Tumor Care

Scientists in South Korea just solved one of cancer treatment's most frustrating puzzles: why the same brain tumor drug works for one patient but fails another.

A team led by Professor Song Ih Ahn at KAIST developed a tiny microfluidic chip that grows a patient's actual tumor cells alongside the blood vessels that surround them. This living replica lets doctors test different treatments and predict which one will work best, all before a patient takes a single dose.

Glioblastoma is among the deadliest brain cancers because it spreads rapidly and varies dramatically from person to person. The brain's protective barrier makes treatment even harder by blocking many cancer drugs from reaching tumors. When glioblastoma forms, this barrier changes in ways that differ for each patient.

Current predictions rely on genetic testing and biomarkers, but those tools miss a crucial piece of the puzzle. They can't capture how a patient's unique blood vessel environment will affect whether drugs actually reach the tumor.

The research team built their chip to include both tumor cells and the vascular barrier that drugs must cross. They cultured patient-derived glioblastoma cells with brain blood vessel cells and supporting tissue inside the microchip. The design even accommodates immune cells, creating a remarkably accurate recreation of where tumor meets healthy brain.

When the team tested chips made from three different patients, the results surprised them. All three patients had identical results on standard genetic testing and were expected to respond the same way to treatment. But on the chips, each patient's vascular barrier behaved differently, and the drugs worked differently too.

The chip predictions matched what actually happened to those patients in the clinic. Same genetics, different outcomes, all explained by the state of each person's blood vessels.

The Bright Side

This platform does something no test could do before: it shows not just whether cancer cells respond to a drug, but how that drug performs in a real tumor environment. Doctors could one day test multiple treatments on a chip made from a patient's own cells and choose the most promising option before starting therapy.

The same technology could speed up development of new cancer drugs by providing a realistic testing ground. Researchers are working to validate the platform with larger groups of patients and refine it for clinical use.

The team envisions a future where personalized treatment planning becomes standard, where patients avoid ineffective treatments and get the right drug the first time.

For families facing glioblastoma's devastating timeline, a test that identifies the best treatment quickly could mean precious months of life gained.