Rutgers Tool Separates Real Tumor Microbes From Lab Error

Scientists at Rutgers Cancer Institute just solved a mystery that's been driving cancer researchers crazy for years.

When doctors sequence tumor DNA, they often find traces of bacteria, viruses and fungi mixed in. These microorganisms could hold the key to why some tumors grow faster, dodge the immune system or resist treatment. But here's the problem: nobody could tell if these microbes actually lived in the tumors or just hitchhiked in from the lab.

Different research teams analyzing the same data reached wildly different conclusions. Some studies claimed tumors throughout the body were packed with microbes. Others called it all contamination.

"There are microbes all over the environment, on our skin and in our breath," said Dr. Subhajyoti De, senior author of the study published in Cancer Cell. "There could be DNA particles floating in the air. How do you know what you're finding came from the tissue you were interested in, or was something introduced along the way?"

The answer is PRISM, which stands for Precise Identification of Species of the Microbiome. This computational tool acts like a super detective, using machine learning to predict whether each detected microbe is real or a contaminant. When tested on samples with known microbial compositions, PRISM achieved accuracy above 90% and beat five other methods.

The researchers put PRISM to work analyzing nearly 4,400 tumor samples across 25 cancer types. The results made much more sense than previous studies. Microbial signals showed up strongest in cancers of the mouth, gut and cervix, areas naturally rich in bacteria. Internal organs showed minimal microbes, exactly what you'd expect.

PRISM also revealed why earlier studies got it wrong. Many "detected" microbes in unexpected places turned out to be common lab contaminants, leaving signatures of how samples were processed rather than what was actually in the tumors.

The real excitement comes from what PRISM found in pancreatic cancer. A subset of tumors contained E. coli that produces a DNA damaging toxin. These same tumors showed molecular changes linked to the dense, fibrotic tissue that blocks drugs and immune cells from reaching pancreatic tumors. Patients with more smoking history tended to have higher microbial levels.

Why This Inspires

The best part? PRISM works on existing data from thousands of patient samples already collected and sequenced. Researchers can now mine this treasure trove without expensive new lab work. "Using standard sequencing that is done to identify the human genome or RNA sequences, as we are doing with PRISM, can obtain this at no additional cost," said lead author Bassel Ghaddar.

This means doctors could soon identify which patients might benefit from microbiome targeted therapies or explain why certain treatments work better for some people. The tool is freely available to academic researchers and could be applied beyond cancer to other diseases.

Real answers about the tumor microbiome could finally lead to real treatments tailored to each patient's unique biology.