Redwing,
As I mentioned in the write-up, it's because there are so few deaths in the median 6.5 years of follow-up. This is especially true since PSA screening caught PC much earlier. It's possible that with longer f/u, we'd see overall mortality differences, but because the median age was 68, we are more likely to see other-cause mortality predominate.
That's why clinical trials on newly diagnosed men with curable disease always use surrogate endpoints. In this case, from the start, they set "biochemical progression-free survival" as the primary endpoint. If we want to look at mortality among high-risk patients, we have to look at retrospective data, which can include large enough datasets to allow for more deaths to have occurred. That's where analyses like the Kishan study become especially useful.
/pcnrv.blogspot.com/2017/02/for-very-high-risk-patients-ebrt-bt-is.htmlYou are right that 2 years of ADT would probably have had better outcomes than 6 months for the high risk guys treated with EBRT only (that was proved by DART 01/05). This was not a test of the usefulness of ADT, so everyone got the same treatment. One can raise the question whether anyone getting the boost therapy actually needed the ADT at all, or whether it was overkill - which is what Dr. Demanes observed on his high risk patients treated with HDR boost therapy. Keep in mind that all patients here received LDR brachy (seeds), so one of the purposes of the long pre-treatment with ADT was to shrink the prostates so that all patients had prostate sizes that were optimized. LDRBT outcomes are poor for men with large prostates. For the EBRT-only cohort, it may have decreased the dose to organs at risk and made their toxicity lower than would ordinarily be expected.
The risk categories used were the standard NCCN risk categories.