PREVAIL Model for Prostate Cancer Survival

Why did you develop the PREVAIL Model for Prostate Cancer? Was there a particular clinical experience or patient encounter that inspired you to create this tool for clinicians?

We desired a prognostic model for contemporary oncology practice, where men are commonly treated with a potent androgen receptor inhibitor such as enzalutamide prior to the use of docetaxel in the mCRPC treatment setting. This model was developed and validated using the phase 3 global PREVAIL study, and data can reasonably be extrapolated to any androgen receptor inhibitor in this setting, pending external validation. 

What pearls, pitfalls and/or tips do you have for users of the PREVAIL Model for Prostate Cancer? Do you know of cases when it has been applied, interpreted, or used inappropriately?

This model is only helpful if all of the variables are measured in the clinic, and we tried to include only those variables that are known to be valid predictors of overall survival in this setting and which are easy to collect in routine practice. External validation is planned as is an updated model with 5 year survival in the coming year. All results should be interpreted cautiously as results may be different based on prior therapies, differences in future therapies after enzalutamide, and the accuracy of this model is only about 70%, meaning that overestimates and underestimates of survival are possible.

What recommendations do you have for doctors once they have applied the PREVAIL Model for Prostate Cancer? Are there any adjustments or updates you would make to the score based on new data or practice changes?

This model can provide a communication tool for patients and providers around prognosis and realistic expectations. This does NOT predict whether enzalutamide will work, and enzalutamide provided a survival benefit irrespective of these prognostic factors and is actually one of the variables used. Results can be updated based on PSA declines post-treatment (Armstrong 2019) as well.

How do you use the PREVAIL Model for Prostate Cancer in your own clinical practice? Can you give an example of a scenario in which you use it?

I use this to estimate survival and communicate this to patients, but I also update these prognostic estimates over time based on an individual patient's response to therapy, which can clearly improve prognosis. I give appropriate caveats given that the model is not perfect, individual outcomes do change and medical practice continues to improve with research advances. But this helps provide a communication framework to inform on treatment decisions and expectations.

Any other research in the pipeline that you’re particularly excited about?

Yes, we are working to update this model based on 5 year data and to externally validate this model in separate large datasets. We constantly strive to update these models as medical practice and treatments improve based on research.

Andrew J. Armstrong, MD, is an oncologist at the Duke Cancer Institute (DCI) in Durham, NC. He is a professor of medicine and surgery, and an associate professor in pharmacology and cancer biology at the Duke University School of Medicine, as well as the director of research in the DCI's Center for Prostate and Urologic Cancer. Dr. Armstrong’s primary research is focused on drug development, biomarkers, and assessing the efficacy of new therapies, as well as improving survival prediction and prognostication in metastatic prostate cancer.

Tomasz M. Beer, MD, FACP, is a medical oncologist and professor of medicine in the division of hematology/medical oncology at Oregon Health and Science University (OHSU). His primary research interests are clinical trials, preclinical investigation, and risk factors in prostate cancer. Dr. Beer is co-lead of the Beer and Qian Lab of the OHSU Knight Cancer Institute.