A new artificial intelligence (AI) tool can help doctors decide how long men with high-risk prostate cancer should stay on hormone therapy. About one-third of patients were found by the AI to not need long-term hormone therapy. These patients could safely stop treatment early without increasing the risk of the cancer spreading.
What the Study Found:
- The AI tool, called the MMAI Prostate LT-ADT Predictive Model, was built using digital biopsy images and data from six large clinical trials.
- It was tested in a seventh trial with over 1,100 men who had high-risk prostate cancer.
- Patients received either short-term (4 months) or long-term (28 months) hormone therapy along with radiation.
- The main goal was to see how many patients developed distant metastasis (when cancer spreads far from the original site).
Key Results:
- Long-term therapy reduced the risk of the cancer spreading in patients who were “biomarker-positive” (meaning the AI predicted they would benefit).
- These patients had a 14% lower chance of cancer spreading over 15 years.
- But for “biomarker-negative” patients, the longer therapy made no difference—meaning they could avoid it and its side effects.
- Long-term therapy also reduced the chance of dying from spread cancer.
Why It Matters:
The AI tool helps doctors know who really needs longer treatment and who doesn’t. This means fewer patients may have to deal with the side effects of long-term hormone therapy, like tiredness, hot flashes, and lower quality of life.
Study Lead and Source:
The research was led by Dr. Andrew J. Armstrong from Duke University and published in the Journal of Clinical Oncology.
Things to Keep in Mind:
- The AI needs clear digital images and full medical records to work accurately.
- Some patients’ data came from different types of procedures, which may affect results.
- The AI still has some limitations, especially if the data it learns from has any bias.
Source: medscape
