AI-based model enhances evaluation in prostate cancer prognostication

AI-Based Model Enhances Evaluation in Prostate Cancer Prognostication

In a landmark study published on 24th October 2023, in JCO Precision Oncology, researchers have demonstrated the efficacy of utilising artificial intelligence (AI) to improve risk classification in patients with localised prostate cancer. The study, led by Dr. Jonathan David Tward from the University of Utah, focuses on developing a sophisticated risk grouping system using multimodal AI (MMAI) models to provide a more accurate prognosis for patients dealing with this prevalent medical condition.

The research encompasses a robust dataset of 9,787 patients who were part of eight phase 3 trials. These patients were treated with various combinations of radiation therapy, androgen deprivation therapy, and chemotherapy. The median follow-up period for assessing these treatments was approximately 7.9 years.

Traditionally, the National Comprehensive Cancer Network (NCCN) risk categories have been used to stratify prostate cancer patients into low, intermediate, and high-risk groups based on certain clinical criteria. In this study, 30.4% of patients were classified as low-risk, 25.5% as intermediate-risk, and 44.1% as high-risk under the NCCN system. However, the introduction of the MMAI risk classification method presented a shift in these numbers, with 43.5% identified as low-risk, 34.6% as intermediate-risk, and 21.8% as high-risk. This innovative approach led to the reclassification of approximately 1,039 patients, which constitutes 42% of the study group.

One of the critical findings was the comparative analysis of 10-year metastasis risks. The metastasis risk for patients within the low-risk category under NCCN was reported at 1.7%, whereas the MMAI method, despite encompassing a broader spectrum of patients in the low-risk bracket, reported a marginally higher metastasis risk of 3.2%. Noteworthy is the further refinement within the high-risk NCCN group by MMAI, which precisely stratified patients into low-, intermediate-, and high-risk groups, with corresponding metastasis risks of 3.4%, 8.2%, and 26.3%, respectively.

The researchers suggest this AI-driven stratification approach is pivotal in balancing treatment strategies, thereby potentially mitigating the risks of overtreatment and undertreatment. This, in turn, enhances the decision-making process involving patients and healthcare providers, allowing them to collaboratively determine the most suitable course of action.

The study does note that several authors have disclosed affiliations with pharmaceutical and biotechnology companies, a common practice that suggests ongoing collaborations within the industry.

Dr. Tward’s team published these findings as part of a broader inquiry into the applications of AI in medical prognostication, highlighting the growing intersection between advanced technology and healthcare. This study marks a significant step towards integrating AI into clinical settings, improving accuracy in risk assessment, and customising treatment pathways for prostate cancer patients.

Source: Noah Wire Services