Skip to main content Skip to secondary navigation
Main content start

New hope from an old drug for patients with aggressive brain cancer

ChEM-H researchers have identified that Prozac could be used in combination with the standard-of-care drug to increase survival among glioblastoma patients.

Glioblastomas are among the most aggressive, hardest to treat brain cancers. Now, new research led by Paul Mischel of Stanford University has shown that an already FDA-approved drug could extend the life expectancy of glioblastoma patients.

“These patients have nothing – no good new treatments in over 30 years. They die usually within one to one-and-half years of diagnosis and they suffer terribly along the way. It is a nightmare,” said Mischel, who is a professor of pathology and an Institute Scholar at Stanford ChEM-H.

Headshot shows Paul Mischel, smiling, with light purple shirt against white background.
Paul Mischel. Image credit: Norm Cyr.

A multi-institution team led by Mischel found that fluoxetine, a drug often sold under the recognizable brand name Prozac, when used in combination with the current standard-of-care drug for glioblastoma, could result in an extended survival of glioblastoma patients. The finding, published on Nov. 2 in Cell Reports, is based on both successful treatment of glioblastomas in mice and on analysis of existing patient insurance data. The researchers are hopeful that the discovery could lead to a new standard treatment—and new hope—for patients.

To find a drug, the research team first had to find a vulnerability. Previous studies from other research groups had given them a hint: glioblastoma cells have different cell membranes than their healthy counterparts. Cell membranes are primarily made up of lipids, fatty molecules that provide structure to isolate a cell’s insides from everything outside a cell. Dispersed among those lipids are other molecules responsible for sending and receiving signals—signals that, among other effects, encourage these cells to grow faster or help them avoid immune detection. Since the lipid composition of glioblastoma membranes is different than that of healthy cells, the researchers reasoned that if they could prevent the cells from making one of the membrane lipids, they might stop the cell from sending signals or even kill these cells altogether, without harming healthy ones.

They turned to existing research papers to find any safe drugs that could both penetrate the brain and interrupt the machinery responsible for making a key membrane lipid. To their surprise, Prozac quickly became a clear candidate.

The team then isolated living glioblastoma cells from patient tumors and, in the lab, treated those cells with Prozac. At doses that left healthy cells unbothered, the tumor cells died. Prozac was killing glioblastoma.

To get a better idea of whether the drug would work on tumors instead of just cells in a dish, the researchers treated mice that had patient-derived glioblastoma tumors with a combination of Prozac and TMZ (temozolomide), the current stand-of-care drug for glioblastomas. Those mice lived longer and had slower-growing tumors than mice treated with TMZ alone.

This led them to a bigger question: why did Prozac work so well? Prozac is an approved selective-serotonin reuptake inhibitor (SSRI), an antidepressant that works by blocking our brain cells’ ability to reabsorb serotonin, a hormone that, when allowed to linger in the spaces between cells for longer, increases the “feel good” signals in the brain. But was this serotonin action also somehow leading to cancer cell death?

Prozac, it turns out, is killing glioblastoma cells through a completely unrelated mechanism. They figured out that the drug prevents cells from receiving signals that tell them to multiply and form new cells, the very signals that give these cancers their aggressive, fast-growing behavior. Prozac interrupts the cell’s ability to make a membrane lipid and, in the process, changes the architecture of the membrane regions that contain these signal-receiving centers. With their altered membranes and interrupted signals, these cells die.

“This was really exciting news for us, but there is a huge difference between a mouse and a human. We needed more evidence that this drug would work in patients,” said Junfeng Bi, the first author on the study and a research scientist in Mischel’s lab.

One of the benefits of repurposing an existing FDA-approved drug is that the researchers already know that the drug is safe for humans. Another benefit is that they could look at medical records to compare the outcomes of glioblastoma patients who were taking Prozac to those who weren’t. Using the IBM MarketScan database of insurance records, they found that Prozac patients taking both Prozac and TMZ had a 50% increase in survival time compared to those who took only TMZ, from under 11 months to over a year and a half.

“With real world data, there are always caveats – these only come from patients who die in the hospital, while most patients are at home, for instance. But the data are remarkably strong,” said Mischel.

Even with the caveats, the team believes this is an important step forward in improving patient care. The researchers have begun planning a potential clinical trial to get closer to an approved therapy.

“Patients and their doctors will be very eager to know what to do. We felt we had to put these data out in the world, and we anticipate it could be practice changing. We will find out,” said Mischel.

The National Brain Tumor Society (NBTS), which also helped fund this study, offers a Personalized Support and Navigation service staffed by a team with deep experience navigating brain tumor treatment options and questions. Patients and families who have questions about this or other studies can contact the NBTS team at


Mischel is the vice chair for research in the department of pathology. Co-authors include Junfeng Bi, Atif Khan, Jun Tang, Aaron M Armando, Sihun Wu, Wei Zhang, Ryan C. Gimple, Alex Reed, Hui Jing, Tomoyuki Koga, Ivy Tsz-Lo Wong, Yuchao Gu, Shunichiro Miki, Huijun Yang, Briana Prager, Ellis J. Curtis, Derek A. Wainwright, Frank B. Furnari, Jeremy N. Rich, Timothy F. Cloughesy, Harley I. Kornblum, Oswald Quehenberger, Andrey Rzhetsky, and Benjamin F. Cravatt.

The work was supported by the National Institute for Neurological Diseases and Stroke, the Defeat GBM Program of the National Brain Tumor Society, the Ben and Catherine Ivy Foundation, the Sharpe/National Brain Tumor Society Research Program, a Compute for the Cure Aware from the Nvidia Foundation, and a UCSD Neuroscience Microscopy Shared Facility Grant.

More News Topics