AI May Assist in Prostate Cancer Diagnosis, Risk Stratification, and Treatment

Artificial intelligence (AI) tools may impact how prostate cancer care is approached with respect to diagnosing patients, risk stratifying disease, selecting treatment, and potentially even surveillance, according to Wayne G. Brisbane, MD in an interview with CancerNetwork^® during Prostate Cancer Awareness Month 2023

Brisbane, assistant professor of urology at the University of California, Los Angeles (UCLA) Health, said that current efforts at his practice are focusing on the feasibility of analyzing proteins and RNAs in urine excretions through AI programs to detect prostate cancer. Additionally, he stated that he is optimistic about the use of AI algorithms in that they may pave the way for new treatment approaches that reduce the likelihood of adverse effects (AEs).

Transcript:

I tell patients that there are 3 to 4 phases of any cancer. The first is diagnosis. There’s a lot of potential [for AI] in diagnosis in terms of determining who has cancer. Currently, we’re using PSA, which is a blunt instrument; it doesn’t give you the granular detail that you need. At UCLA, we’re looking at the proteins and RNAs [excreted through urine] to see whether those can be [analyzed] with the power of machine learning or AI algorithm to decipher through all those different urine proteins to detect cancer.

The second thing is risk stratification. Once it is determined that a patient has cancer, we need to know how risky it is. Prostate cancer is a very heterogeneous disease. Some men can live with prostate cancer without any risks of their health. There are some men who die from prostate cancer; we need to figure out who is in what bucket so we can treat them appropriately. I think that AI can help with that, as well.

The third is going to be treatment. We’re still figuring out exactly how AI can help with treatment. In radiation oncology, they’re using sophisticated technology to aim their beams. We’re doing some clinical trials in focal therapy, where we’re saying, ‘Okay if we can predict exactly where the tumor is, can we just ablate that tumor?’ AI will open up a new treatment area of trying to just kill the tumor and minimize collateral damage. We have a lot of work to do to figure out if it works. Men who are on those clinical trials are helping us advance the field. I’m very optimistic about it as a [means] of opening up more cancer care with fewer [AEs], which is the goal of all of radiation oncology, urology, and medical oncology.

Finally, there is surveillance. We’ll have to see exactly how AI will help us with surveillance. I think it’ll be important as well.