New Tool Finds Real Bacteria in Tumors, Improves Treatment

Scientists just solved a mystery that's been puzzling cancer researchers for years, and the answer could change how we treat the disease.

When doctors analyze tumor DNA, they often find traces of bacteria, viruses and fungi mixed in with the cancer cells. But here's the problem: nobody knew if those microbes actually lived in the tumors or just contaminated the samples during testing. Different research teams looking at the same data reached completely opposite conclusions.

Now, researchers at Rutgers Cancer Institute developed a tool called PRISM that finally settles the debate. The software can tell the difference between real microbial signals and contamination with over 90% accuracy.

"There are microbes all over the environment, on our skin and in our breath," said Dr. Subhajyoti De, who led the study published in Cancer Cell. "There could be DNA particles floating in the air."

PRISM works like a super-smart filter. It screens tumor DNA samples, removes human genetic material and compares what's left to databases of known microbes. Then a machine learning model predicts which bacteria were truly present in the tumor and which hitchhiked in from lab surfaces, reagents or even air.

The researchers tested PRISM on nearly 4,400 tumor samples across 25 cancer types. The results made much more sense than previous studies. Cancers in microbe-rich areas like the mouth, gut and cervix showed the strongest bacterial signals. Internal organs that don't connect to the outside world had minimal microbes, just as doctors would expect.

In pancreatic cancer, PRISM found something fascinating. Some tumors contained E. coli bacteria that produce a toxin damaging DNA. Those same tumors showed changes in how proteins behaved, potentially explaining why pancreatic tumors build thick, fibrous barriers that block drugs and immune cells from reaching cancer.

The Ripple Effect

This breakthrough means researchers can now mine thousands of existing tumor samples without expensive new lab work. Every major cancer study that's already been done becomes a potential goldmine of information.

The discovery could reveal why certain treatments work brilliantly for some patients but fail for others. It might identify which people would benefit from therapies targeting their tumor's microbiome. Doctors could potentially prescribe treatments based on which bacteria live in a patient's specific cancer.

"We can generate hypotheses about which patients might respond to certain therapies or which microbial metabolites might be druggable targets," said lead author Bassel Ghaddar.

The tool is freely available to academic researchers, meaning labs worldwide can start using it immediately to unlock insights from data they've already collected. No waiting, no additional costs, just answers that were hiding in plain sight all along.

After years of confusion and conflicting results, cancer research just got a whole lot clearer.