According to ScienceAlert, researchers from UC San Francisco and Gladstone Institutes discovered that two FDA-approved cancer drugs—letrozole and irinotecan—significantly reversed Alzheimer’s brain damage in mouse models when used together. The 2025 study used computational tools to analyze the Connectivity Map database and cross-reference patient records, finding these drugs decreased Alzheimer’s risk in cancer patients. When tested in Alzheimer’s mice, the combination reduced harmful tau protein clumps and improved learning and memory capabilities. The approach targets different brain cells, with letrozole working on neurons and irinotecan on glia cells. With over 55 million people currently living with Alzheimer’s worldwide—a number expected to more than double in the next 25 years—the findings offer potential for repurposing existing medications.
The computational breakthrough
Here’s what’s really interesting about this approach. Instead of the traditional “one drug, one target” method that’s failed repeatedly in Alzheimer’s research, the team used computational biology to analyze complex gene expression patterns. They looked at the Connectivity Map database to find drugs that could reverse the specific gene changes seen in Alzheimer’s. Basically, they let the data point them toward solutions rather than starting with preconceived notions about what should work. Computational biologist Marina Sirota said their tools “opened up the possibility of tackling the complexity directly”—and that’s crucial because Alzheimer’s involves dozens of genes and proteins going wrong simultaneously.
Why this matters
Look, we’ve seen countless Alzheimer’s treatments show promise in mice only to fail in humans. But this approach feels different. The researchers didn’t just pick random drugs—they found medications that had already shown reduced Alzheimer’s risk in cancer patients taking them. That’s real-world data suggesting the effect might translate to humans. Plus, the combination therapy approach makes biological sense. Alzheimer’s isn’t one problem—it’s multiple systems failing. Using two drugs that target different cell types (neurons and glia) acknowledges that complexity. Neuroscientist Yadong Huang nailed it when he said traditional drug development struggles because it tries to fix Alzheimer’s with “one drug for a single gene or protein that drives disease.” This is thinking bigger.
The reality check
Now for the sobering part. These are cancer drugs with significant side effects. Letrozole can cause bone pain, hot flashes, and joint stiffness—not exactly mild symptoms for elderly patients. Irinotecan comes with its own set of challenges including diarrhea and bone marrow suppression. The question becomes: would Alzheimer’s patients tolerate these side effects for potential cognitive benefits? And we’re talking about mice here. The jump from mouse models to human Alzheimer’s is enormous—most treatments that work in mice fail in people. The researchers acknowledge this and say clinical trials are the necessary next step. But given that Alzheimer’s numbers are exploding globally, the potential reward might justify the risk.
What comes next
So where does this go from here? The researchers are hopeful about swift translation to human trials, but let’s be real—drug repurposing still requires extensive testing. The beauty is these drugs are already FDA-approved, which could potentially shorten the timeline. But we need to see if the cognitive improvements in mice translate to humans, and whether the side effect profile is manageable for Alzheimer’s patients. The personalized medicine angle is intriguing too—matching treatments to individual gene expression patterns could be the future. As Sirota said, “If completely independent data sources… guide us to the same pathways and the same drugs, and then resolve Alzheimer’s in a genetic model, then maybe we’re on to something.” After decades of failed Alzheimer’s trials, that “maybe” feels more substantial than usual.
