A study led by UCLA investigators shows that artificial intelligence (AI) could play a key role in improving treatment outcomes for men with prostate cancer by helping physicians determine who is most likely to benefit from partial gland cryoablation, a minimally invasive procedure that treats localized prostate tumors.
Researchers found that an AI tool called Unfold AI, developed by researchers at UCLA and Avenda Health, accurately estimates prostate tumor volume and helps identify patients with a higher chance of successful treatment.
The study, published in BJUI Compass, suggests that using AI to measure tumor size could reduce treatment failures by more than 70%.
“By using AI to measure the size of a man’s prostate tumor more precisely, we can better predict who is likely to be cured with focal therapies like partial gland cryoablation,” said Dr. Wayne Brisbane, assistant professor of urology at the David Geffen School of Medicine at UCLA, member of the UCLA Health Jonsson Comprehensive Cancer Center, and first author of the study.
Partial gland cryoablation freezes and destroys only the cancerous part of the prostate, instead of removing or treating the entire gland. The approach eliminates cancer cells while minimizing damage to vital areas, resulting in fewer side effects than surgery or radiation and offering patients a better quality of life. The technique involves using imaging guidance, such as MRI, to accurately locate the tumor and guide the treatment. Real-time imaging during the procedure helps monitor treatment progress and ensures the precise delivery of energy to the intended area.
Current methods, however, tend to underestimate how big the tumor really is, and they can miss smaller cancer spots, which can lead to incomplete treatment and cancer recurrence.
Unfold AI helps with this issue by analyzing data from MRI scans and biopsies to create a detailed, three-dimensional map of the prostate tumor. This helps doctors more accurately see the true size and boundaries of the cancer.
To evaluate the accuracy of the AI software, the team enrolled 204 men with localized prostate cancer who underwent partial gland cryoablation in a clinical trial at UCLA between 2017 and 2022. All participants received MRI-guided biopsies at diagnosis, as well as follow-up biopsies at 6 and 18 months after treatment to monitor for cancer recurrence.
In the trial, physicians used Unfold AI to generate a 3D map of each tumor, estimating its true volume. They compared tumor volume to traditional indicators such as tumor grade and PSA levels to predict treatment outcomes.
They found that tumor volume was the strongest predictor of treatment success, while tumor grade did not correlate a successful outcome. Patients with tumors smaller than 1.5 cubic centimeters had significantly better outcomes after cryotherapy and these men were less likely to need further treatment or develop metastases. Using this tumor volume threshold as an eligibility criterion would have prevented 72% of treatment failures, according to the study.
“With Unfold AI, doctors now have a method to determine the volume of cancer within a prostate tumor,” said Dr. Leonard Marks, professor and deKernion Endowed Chair in Urology at the David Geffen School of Medicine at UCLA, member of the UCLA Health Jonsson Comprehensive Cancer Center, and senior author of the study. “Such a method has not been previously available. It's important because tumor volume is a major determinant of treatment success or failure. Using AI to predict tumor volume and shape gives a clearer picture and could help choose better candidates for focal cryotherapy.”
While these findings are promising, the researchers emphasize the need for larger, multi-center trials to confirm their results.
“The study marks an important advance in integrating AI into prostate cancer treatment decision-making, offering the potential for more personalized prostate cancer care,” said Brisbane.
The collaborative research at UCLA and the UCLA-backed startup Avenda Health that led to the development of Unfold AI was supported by grants from the National Institutes of Health.
Other authors include Alan Priester, Anissa Nguyen, Mark Topoozian, Merdie Delfin, Samantha Gonzalez, Kyla Grunden, Shannon Richardson and Shyan Natarajan from UCLA and Sakina Mota from Avenda Health.
Marks and Natarajan are cofounders of Avenda Health.
Journal
BJUI Compass