Microfluidic Chip Reveals How Living Glioblastoma Slices Resist Chemotherapy

Scientists at Columbia University have developed a microfluidic chip that enables detailed study of how living slices of human glioblastoma tumors resist chemotherapy, offering new insights that could guide the development of more effective treatments for this aggressive brain cancer.

The system, described in research published in the journal Lab on a Chip, uses a type of microfluidic device to maintain and analyze tumor tissue samples obtained directly from patients during surgery. These living slices preserve the complex cellular architecture of the tumor, allowing researchers to observe how different cell types within the tumor respond to drugs in real time.

According to Peter Sims, PhD, associate professor of systems biology at Columbia and senior author on the study, the approach builds on earlier work where tumor slices were grown in petri dishes to test drug effects. While those experiments provided valuable information about how individual cell types respond to treatment, Sims sought to improve the system by automating tissue maintenance and obtaining more quantitative data.

The breakthrough came through collaboration with Sam Sia’s lab, located across the hall at Columbia’s Irving Medical Center. After connecting at a research symposium, the teams combined microchip engineering techniques with advanced gene profiling to create a perfusable microfluidic chip capable of supporting parallel culture and drug testing on multiple thick tissue slices from human glioblastoma resections.

The chip design allows fluids to flow through the device, perfusing the tumor slices with nutrients and drugs while enabling continuous monitoring. This setup supports single-cell screening of anti-cancer compounds, revealing how various tumor cell populations survive chemotherapy exposure. Early findings using the system have already identified specific mechanisms by which glioblastoma resists treatment, including adaptive changes in certain cell subpopulations that persist despite drug exposure.

By maintaining the tumor’s native microenvironment outside the body, the microfluidic platform overcomes limitations of traditional cell cultures, which often fail to reflect the complex interactions between tumor cells, blood vessels, and surrounding tissue seen in actual tumors. The ability to sequence individual cells from the slices after drug treatment provides a high-resolution view of molecular shifts associated with resistance.

Researchers note that glioblastoma remains one of the most difficult cancers to treat, with a median survival of approximately 14 months and a recurrence rate exceeding 90% following initial therapy. Standard treatments typically involve surgery followed by radiation and chemotherapy with temozolomide, but tumor cells frequently develop ways to evade these interventions.

The microfluidic chip approach represents a step toward personalized medicine strategies, where a patient’s own tumor tissue could be tested outside the body to predict which drugs are most likely to be effective before administering treatment. While the technology is still primarily used in research settings, scientists believe it holds promise for improving preclinical drug evaluation and identifying novel therapeutic targets.

Ongoing work focuses on refining the chip’s design to increase throughput and compatibility with additional analytical methods, such as imaging and proteomics. The team also aims to expand testing to include immunotherapies and combination regimens, which are increasingly being explored for glioblastoma but have shown limited success in clinical trials to date.

As of April 2026, the microfluidic chip system has been used to study tumor samples from multiple patients, generating data that researchers say could help explain why certain therapies fail and point toward more precise intervention strategies. The study was conducted at Columbia University Irving Medical Center and supported by institutional resources, though specific funding details were not included in the reported findings.