AI Tool Offers Prognosis for Patients with Head and Neck Cancer

Current therapies often involve combinations of surgery, radiation, and chemotherapy, which can be difficult to tolerate and may leave lasting side effects. A major challenge for clinicians is identifying which patients are likely to experience aggressive disease spread before treatment begins. Researchers have now developed a noninvasive artificial intelligence (AI)-based approach that can predict the likelihood of cancer spread using routine imaging, helping guide treatment intensity decisions earlier in care.

In research led by Mass General Brigham (Boston, MA, USA), in collaboration with Dana-Farber Cancer Institute (Boston, MA, USA), the team developed an AI model designed to analyze standard computed tomography (CT) scans obtained during routine cancer evaluation. The AI tool focuses on detecting pathologic extranodal extension (ENE), a condition in which cancer cells spread beyond lymph nodes into surrounding tissue. ENE is a critical prognostic marker, but it is currently confirmed only after surgical removal and examination of lymph nodes, limiting its usefulness in pre-treatment planning.

Using CT imaging data, the AI model predicts the number of lymph nodes affected by ENE before any surgical intervention. This information provides an early indicator of how aggressive the cancer may be and whether a patient is likely to benefit from intensified therapy or could safely undergo less intensive treatment. By integrating imaging-based predictions with existing clinical risk factors, the tool offers a more refined and personalized assessment of disease behavior without requiring invasive procedures.

The researchers evaluated the AI tool using CT scans from 1,733 patients with oropharyngeal carcinoma. The model successfully identified patients at higher risk of uncontrolled cancer spread and poorer survival outcomes. When the AI-derived ENE predictions were combined with established clinical staging systems, overall risk stratification improved, allowing for more accurate predictions of patient prognosis.

The findings, published in the Journal of Clinical Oncology, could help clinicians determine which patients may benefit from aggressive multimodal treatments such as combined chemotherapy, radiation, or immunotherapy. Conversely, it may also identify patients suitable for treatment de-intensification, such as surgery alone, reducing unnecessary toxicity. Going forward, the researchers suggest that this AI-based biomarker could be incorporated into future staging systems and clinical decision-making workflows, as well as used to select patients for clinical trials evaluating novel treatment strategies.

“Our tool may help identify which patients should receive multiple interventions or would be ideal candidates for clinical trials of intensive strategies such as immunotherapy or additional chemotherapy,” said senior author Benjamin Kann, MD. “Our tool can also help identify which patients should undergo de-intensification of treatment, such as surgery alone.”

Related Links:
Mass General Brigham
Dana-Farber Cancer Institute