Medical imaging could improve precision diagnosis of aggressive head and neck tumors

Medical imaging routinely used in cancer care may hold far more biological information than previously thought. An international study involving Umeå University guest professor Lukas Kenner shows that PET/CT scans can capture the molecular activity of particularly aggressive head and neck tumors, opening new possibilities for more precise diagnosis and treatment planning.

Most head and neck cancers are so-called squamous cell carcinoma, that affect tissues of the mouth, throat and larynx. tumors that are HPV-negative – that is, not caused by the human papillomavirus, HPV – tend to be aggressive and difficult to treat. Reliable markers that predict how these cancers will progress, and how they might respond to therapy, have been lacking. However, a new study suggests that medical imaging already in use may also be useful for identifying and monitoring high‑risk tumors.

"Imaging has traditionally been used to determine where and how large a tumor is," says Lukas Kenner, guest professor at Umeå University and corresponding author. "Our results show that the images also contain information about the tumor's molecular behaviour, meaning how aggressive it is."

Normally, the aggressiveness of a tumor would be determined through biopsies, which are small tissue samples from the tumor that are examined with a microscope. However, biopsies are invasive, time-consuming, and only sample small parts of the tumor, so the most aggressive regions can sometimes be missed. In this study, the research team analysed tumor samples together with clinical PET/CT imaging from patients with HPV-negative head and neck cancer. They found that tumors driven by activation of the Hedgehog signalling pathway, a key regulator of cancer growth, exhibit distinct patterns on PET/CT scans. These tumors accumulate more of the tracer used in imaging, reflecting higher biological activity and poorer prognosis. To test whether imaging also reflects treatment effects, the researchers blocked the Hedgehog pathway in laboratory experiments. Tumor growth slowed, and the PET/CT signal changed in parallel, mirroring the patterns seen in patients in the clinic.

"This is particularly encouraging because it suggests we may be able to monitor biological treatment response non-invasively," Kenner says.

By combining molecular multi-omics with quantitative imaging analysis, the study establishes a framework linking tumor biology to radiological features. Such approaches are expected to contribute to AI-supported cancer diagnostics and more individualised therapy decisions.

"Before this can be used clinically, prospective studies are needed," Kenner notes. "But the work shows how integrating imaging and molecular data can move precision oncology forward."

Source:

Umea University

Journal reference: