Analytics in Action News

Artificial Intelligence Can Predict Prostate Cancer Recurrence

An artificial intelligence tool is able to examine data from MRI scans and predict the likelihood that prostate cancer will recur after surgical treatment.

Artificial intelligence can predict prostate cancer recurrence

Source: Getty Images

By Jessica Kent

- Artificial intelligence can analyze pre-operative MRI scans to predict patient outcomes following prostate cancer surgery, a study published in EBioMedicine.

A critical factor in managing prostate cancer in men undergoing surgery is identifying which are at highest risk of recurrence and prostate cancer-specific mortality. Researchers noted that approximately 20 to 40 percent of patients experience recurrence and may develop further metastasis after definitive treatment.

Previous efforts proposed several tools to identify patients at high risk of prostate cancer recurrence after surgery, including both pre-operative and post-operative tools. However, the performance of these tools varies between different cohorts.

Researchers developed and evaluated an artificial intelligence tool to examine a range of data, including pre-operative MRI scans and molecular information. The team collected MRI scans from Cleveland Clinic, Mount Sinai Hospital, University Hospitals, and the Hospital of the University of Pennsylvania to validate the algorithms.

Researchers applied their tool, called RadClip, to pre-operative scans from nearly 200 patients whose surgeons removed their prostate gland because of cancer. The group then compared the tool’s results to those of other predictive approaches, as well as patients’ outcomes in the subsequent years.

The AI algorithms were able to accurately identify subtle differences in heterogeneity and texture patterns inside and outside the tumor region on pre-operative MRI to predict patient outcomes after surgery.

When compared to other pre-operative tools like the Prostate Cancer Risk Assessment (CAPRA) score, RadClip achieved an area under the curve (AUC) of 0.71 while CAPRA reached an AUC of 0.69. Decipher, the most utilized test in real-world practice, yielded an AUC of just 0.66.

These results demonstrate the ability for AI algorithms to accurately predict the risk of disease recurrence.

“This tool can help urologists, oncologists and surgeons create better treatment plans so that their patients can have the most precise treatment,” said Lin Li, a doctoral student in Case Western Reserve’s Biomedical Engineering Department and a member of the Center for Computational Imaging and Personalized Diagnostics (CCIPD) team that developed the tool.

“RadClip allows physicians to evaluate the aggressiveness of the cancer and the response to treatment so they don’t overtreat or undertreat the patient.” 

While these findings are promising, researchers noted that clinical trials will need to demonstrate that the tool can also help identify men undergoing surgery who would also benefit from additional therapy.

The approach was unique in that it used a range of data to predict patient outcomes, the team stated.

“We’re bringing together and connecting a variety of information, from radiologic scans like MRI to digitized pathology specimen slides and genomic data, for providing a more comprehensive characterization of the disease,” said Anant Madabhushi, CCIPD director, Donnell Institute Professor of Biomedical Engineering at Case Western Reserve and the study’s senior author. 

The study also demonstrates the value of imaging data, showing that RadClip provides better prognostic information than other commonly used tools.

“Genomic-based tests cost several thousand dollars and involve destructive testing of the tissue,” Madabhushi said. “Prognostic predictions from an MRI scan provide a non-invasive method for making both short-term and long-term decisions on treatment.” 

Researchers can use data generated from the AI algorithms to address two important clinical areas: prostate surgery and post-operative management. Additionally, information gathered from pre-operative MRI images can help predict the existence and extent of cancer on the margins of tumors, which would allow surgeons to make informed decisions about how much tissue to remove.

Data can also predict the risk of cancer recurrence so oncologists can determine whether a patient needs adjuvant treatments after surgery, like radiation therapy or chemotherapy.

“Having this information before surgery provides surgeons and oncologists the time and space to adjust treatment plans and come up with a plan that’s best suited to the patient,” Li concluded.