Bold claim: Daily radiotherapy scans could double as early warnings for side effects, helping doctors tailor treatment and spare patients from rectal bleeding. But here’s where it gets controversial: the same imaging already collected for beam setup might unlock a future of adaptive radiotherapy without asking for new scans or equipment.
Researchers analyzed daily imaging from 187 men undergoing prostate cancer radiotherapy to see if machine-learning could link radiomic features—subtle texture and pattern signals in the images—to later toxicity within two years. They found that signals visible in the first week of treatment were surprisingly predictive of subsequent rectal bleeding. When data from the first three weeks were combined, predictions became even more reliable. This suggests an early, actionable window to adjust radiotherapy—potentially reducing side effects while preserving cancer control.
In practical terms, the study implies that the information already captured during daily setup imaging could inform adaptive treatment decisions. If further validated, clinicians might monitor radiomic changes to decide when and how to modify plans mid-course, enhancing personalization of therapy.
The work was published in Physics and Imaging in Radiation Oncology. It was funded by Prostate Cancer UK, with contributors from the University of Cambridge and The Christie NHS Foundation Trust.
Experts emphasize that this is a promising proof-of-concept rather than a ready-to-implement protocol. Larger trials and automation will be essential before such methods influence routine clinical decisions. Still, the finding points toward a kinder, smarter approach to radiotherapy—one that leverages existing data to forecast toxicity risks and tailor treatment accordingly.
As with any new strategy, opinions vary. Do you think early imaging biomarkers should drive treatment adaptation now, or ought we to wait for more evidence and standardized guidelines? What thresholds would you consider acceptable for altering a proven cancer-control plan? Share your thoughts in the comments.
Source: University of Edinburgh
16.12.2025
