Thanks for sharing that. Note especially this line

"Between 1998 and 2015, 204 patients with HR localized prostate cancer…"

In other words, the participants in the study all had early-stage ("localized") prostate cancer, not advanced prostate cancer.

There's also the risk that findings based on data they started collecting almost 30 years ago (before even 1st-gen ARSIs like Zytiga were available) might no longer apply, especially now that we have significantly more-effective modern treatments like the -lutamides, but that's another discussion. 🙂