"We're pleased to report that there was a strong prognostic signal we found with the overall cohort," says Charles Parker, MD.
In this video, Charles Parker, MD, and Gerhardt Attard, MD, PhD, discuss notable findings from the study “External validation of a digital pathology-based multimodal artificial intelligence (MMAI)-derived model in high-risk localized (M0)/metastatic (M1) prostate cancer (PCa) starting androgen deprivation therapy (ADT) in the docetaxel (Doc) or abiraterone (AAP) phase III STAMPEDE trials,” which was presented at the 2023 European Society for Medical Oncology Annual Congress in Madrid, Spain. Parker is a clinical research fellow and PhD student at UCL Medical College in London, the United Kingdom. Attard is the John Black Charitable Foundation Endowed Chair in Urological Cancer Research at University College London.
Parker: The big unanswered question for us in this study was, could this tool that's been developed in a much earlier stage of disease in patients receiving completely different treatment still have prognostic validity in our patient cohorts in this later, more advanced stage of the disease? And we're pleased to report that there was a strong prognostic signal we found with the overall cohort—a hazard ratio of 1.72 for increased risk of an event of prostate cancer-specific mortality for every standard deviation increase in the score, which fits across both the metastatic and nonmetastatic settings and across all treatment paradigms.
Attard: I'll give Charles a break and answer your question about whether the findings are surprising. And they surely are, because these were data that were lying there, clinical data that we collect as routine, images that we collect as routine. Now, of course, we have scanned them, so we've done additional things to enable this analysis. But there are bits of information that are collected routinely on every patient diagnosed with this advanced stage of prostate cancer. And each of those individual bits has very low ability to predict whether a patient is going to die from prostate cancer or not. In fact, we do not use them for our prediction, but all pulled together and put together using this model, as Charles explained, we then are able to very accurately predict whether a patient is going to die from prostate cancer or not, which is clearly vital information when both making a decision on what treatment we want to give or combination of treatments when counseling patients. And I think this will increasingly be implemented into clinical practice.
This transcription was edited for clarity.