Read the Full Video Transcript

Daniel Spratt: Now we're talking post-operative; the guy chose to have surgery and he's had a biochemical recurrence. So should we add systemic therapy? What type and for how long? So as you've heard briefly, four of the primary phase III randomized trials have been discussing at least using hormone therapy. And to kind of jump to the punchline—and I'll explain why I conclude this—in general, patients with a lower PSA, I'll say less than 0.5, do not clearly derive overall survival benefits from the addition of hormonal therapy to early salvage radiation. Those receiving late salvage radiation, in general, may improve overall survival from their use of hormone therapy, where the signal is more clear with long-term hormone therapy. The point really is, though, this is a very heterogeneous disease space.

So this got kicked off by the RTOG 9601 trial, which, remember, was salvage radiation—typically late salvage radiation therapy, very late by today's standards—with plus or minus two years of high-dose bicalutamide. And it took 13 years of median follow-up until that survival benefit. And you can see the upper bound of the confidence interval finally became statistically significant.

In the era of this trial, you can see, they stratified patients by PSAs of 1.5. That's how common it was. And if you look by that pre-specified PSA stratum, that 15% of patients receiving very late salvage radiation—these are the patients that derive the survival benefit, 25% absolute improvement. These guys, if you had PSMA PET, probably similar to the EMBARK trial, were likely metastatic. The 85% of the trial that has the PSAs less than 1.5—there is absolutely no improvement in survival, and we can have debates ad nauseam about this, but you can look in this subset, you can look at the GETUG-16 trial long-term follow-up. You can look at the RTOG-SPPORT trial—I'll call it medium follow-up. There is absolutely no signal of a survival benefit in these studies in not even truly early salvage radiation.

So if you look at just an aggregate meta-analysis of those studies, excluding the ultra-high PSA patients, the hazard ratio is 0.98. And there still are many patients receiving salvage radiation at PSAs over 0.5. This again, going back to, this would never pass any regulatory approval, and this is very distinct from localized prostate cancer.

So the big addition here is the RADICALS-HD trial. These three different phase III randomized trials that sort of got combined, as you see here. This was already shown, but of course men choosing, and providers choosing, to put patients on none versus short are far more favorable. There's very little high-grade disease. Most patients had lower PSAs. Those randomized to receive short versus long. There were more patients, while still the minority, with PSAs over 0.5.

Again, now we've got the fourth study and this is a massive study. Each arm on this trial is the size of RTOG 9601 or the GETUG trial. I can't see any difference here in these curves. Then you look at the short versus long-term hormone therapy. The OS is on the right; the MFS, the primary endpoint, is on the left. I will say, and I said this similarly about the GETUG trial, if you look at the difference in testosterone recovery of two years of ADT, which is almost four years in duration, that is nearly the identical delay in the benefit of MFS. So you are delaying testosterone recovery. You are delaying PSA rising and the delay of detecting METs. There's no flattening of these curves saying this truly is resulting in improvement in cure. But this is statistically significant.

However, when you look at the three-way randomization—underpowered unfortunately—but these are probably more favorable patients. When you look at radiation alone versus two years of ADT now, this is the hazard ratio of 0.94. Again, no difference, and this is an MFS.

So, personalizing therapy. So to kind of disagree a little with the last speaker, this is not simply retrospective, institutional data. This is from a phase III randomized trial, lower PSAs versus as the PSA goes up, these patients are metastasizing before you give radiation. So there is a substantially worse all-cause mortality and distant METs if you delay the use of radiation therapy, and studies by many nuclear medicine colleagues—this is just one of the studies combined with that Stephenson nomogram. You can see that the success rate at 10 years of salvage radiation correlates very nicely with the detection of PSMA PET nodal or distant METs. If we're not radiating those disease, of course they're going to recur.

So probably the best way to personalize therapy is the PSA. We showed this in 9601; the primary endpoint was overall survival. There was a statistically significant interaction by, as the PSA goes up, the benefit of adding hormone therapy dramatically goes up. In the GETUG trial, they did not show an interaction test. Its primary endpoint was biological or biochemical progression-free survival. Similarly, PSAs greater than 0.5 had a much greater absolute and relative benefit from an addition of six months of hormone therapy. You look at the SPPORT trial; again, I only can show what's been presented. This is PSAs greater than or less than one. Difference in absolute as well as relative benefits, and you can see at the bottom left, this was presented at a meeting previously, that the five-year distant METs, if you add short-term hormones, is only 2% for low PSAs versus 11% for high PSAs.

RADICALS-HD zero over months. There did include adjuvant versus early salvage patients. The interaction test is nearly significant. Adjuvant patients did not seem to derive a benefit from ADT.

You look at the six versus 24 months. Now you could see here that I'm maybe reaching with this arrow, but you can look at the hazard ratios going from 0.83, 0.69, 0.67, and astute residents will say, "Dan, this disagrees with your paper on 9601." It actually does not, because they did not separate out the high PSA patients in their study. And so, in this intermediate range there's intermediate potential benefit, but at low PSAs at early salvage, which really should be at 0.1 or 0.2, there is no benefit.

In FORMULA-509, the trial that was mentioned of six months of ADT plus or minus AB and APA with salvage radiation. MFS was significantly improved with, again, a near statistical interaction by those with higher PSAs.

So moving beyond PSA, there's of course the Decipher test. This is from a trial we ran. The results are not out, but about half of patients receiving early salvage radiation therapy will have a high genomic classifier score. About half will have a lower score. And as shown in 9601, those with lower Decipher scores have a very low rate of PCSM, in the blue curve, versus higher scores. And the absolute benefit of the two years of bicalutamide—you can see in the low scores, that's basically 0% versus the high scores is over 10%. And this has been shown previously in the SAKK trial showing its prognostic.

So this was briefly shown by Dr. Efstathiou. This will be presented as one of the oral presentations at AUA. So do not—I will find you—put this online. This is a study of 9601 and 0534, two of these phase III trials where we have generated new MMAI models. And again, it is highly prognostic. It was not designed to be predictive. But you can see here that it is very clearly—the signal of being predictive is there of who benefits or not from hormone therapy.

So in summary, hormone therapy, I think, has no clear survival benefit in unselected patients receiving early salvage radiation. This is very distinct from my earlier talk in localized prostate cancer. Remember, ADT is thought to radiosensitize patients; that's the benefit. Dose escalation, as we heard earlier by Neha, improves outcomes in localized disease. Every dose escalation trial has been negative in the post-operative setting.

The signal of greater benefit is in higher PSAs. This correlates strongly with patients that probably have nodal or distant METs. High 22-gene GC, which also has been shown to correlate with PSMA-positive nodal disease. And now soon it'll be the MMAI score.

So thank you so much.