Jonathan Tward: It's my pleasure, Dr. Agarwal. It is a special pleasure to be sitting across from you knowing that you and I were trainees together, and now a good 20 plus years later, here we are talking about some of the most fascinating science in our field.
Neeraj Agarwal: What can be more fun than this?
Jonathan Tward: Indeed.
Neeraj Agarwal: You have done lot of work in patients who have biochemical recurrence after definitive surgery, radical prostatectomy. We just saw the data in ASCO GU 2026 in this meeting from POSEIDON analysis. Patient-level data from more than 5,000 patients from six randomized clinical trials that adding androgen deprivation therapy to salvage radiation therapy in those patients who do not have a PSA 0.5 did not really help them. With a really long follow-up, we saw 10-year follow-up, which is great, but now field has moved on in last 10 years. I think we need to select our patients better. It's great to have those data. Kudos to the investigators. But especially given your work with the molecular classifiers, how to better select these patients. So tell our audience today, worldwide audience, what should we be doing? And especially focusing on your abstract you presented in ASCO GU meeting.
Jonathan Tward: Yeah, thank you. One of the blessings of oncology is that over the past 30, 40 years, we've been able to stratify people into cohorts. That's been helpful because it helps harmonize treatments and make sure at the population level that people get the best possible therapies. The problem is when you treat people at the population level, you are in fact often over-treating or under-treating, so we have a lot of questions with regard to intensification and deintensification. If we go specifically into the post-prostatectomy, biochemically failed space, we have multiple randomized trials, as you alluded to, that have shown quite clearly that if you give androgen deprivation therapy to radiation therapy, you have a tendency to improve progression-free survival, metastasis-free survival. In some cases, it's been reported overall survival. The problem is that those trials were designed and completed many years ago and patients didn't use to be referred to us quite as quickly as they are now. There's been a recognition that you shouldn't wait on people for their PSA to get above very high threshold.
The question kind of becomes, what's relevant in the year 2026? You alluded to an amazing meta-analysis that Dr. Kishan, along with Dr. Spratt and others have worked on, that demonstrated that in spite of these randomized trials, which at the population level show there's a benefit from adding ADT. In truth, there doesn't appear to be a benefit when you use PSA, for instance, as a biomarker. For instance, PSA 0.5. So now the question becomes, should we just wholesale not give anybody androgen deprivation at a PSA level that's under 0.5? I think that would be one interesting approach that you could take, but recently there were other trials, for instance, one by the NRG called GU6, that was designed around a molecular signature called PAM50. It was specifically designed to ask the question, can we actually select patients who we think would have a better benefit of the addition of some kind of hormone therapy? On that trial, they were using a genomic classifier that Veracyte makes, the Decipher test. But as a component to that test, you derive this PAM50 signature. What they showed in that clinical trial is that that signature will stratify patients into one of three groups, what they call luminal B or luminal A or basal. If you separately look at the luminal Bs from the non-luminal Bs, they showed on that trial very convincingly that it is one of the first and only predictive biomarkers that can actually tell you in this patient population who you should use the hormone therapy in, even if the PSA is low. 85% of patients on that trial had PSA of something like 0.5 or less. The question that I wondered about is that it is not considered a standard of care to use an ARI only in the post-prostatectomy space.
We traditionally use androgen deprivation therapy. Although some people muddle the two mechanisms of action, it still in my mind remained an unanswered question, like we show with an ARI, you have a predictive biomarker. But it wasn't 100% clear to me that we could translate it to androgen deprivation therapy. I was able to leverage the power of our prospectively collected database, which is pretty massive. Although it's not a multi-institutional trial, we could look within our own dataset at Huntsman where we had PAM50 signatures on about 130 patients. We were able to show that it appears that effect in luminal B holds with ADT as well, which is in fact the care standard. What's really the big picture? The big picture here is I think based on the POSEIDON meta-analysis and others, if you don't have access to molecular testing, I would withhold ADT in people under a PSA 0.5. But if you do have access to molecular testing, the PAM50 signature, which was shown in a prospective trial to be a stratifier for hormone benefit in the ARI use appears in our dataset to hold up in the androgen deprivation therapy mechanism of action as well.
Neeraj Agarwal: Such a fascinating discussion. Even for me, I learned so much from you just now in the last few minutes. So just to summarize, so first of all, thank you for such a clear explanation of what is going on and what we should be doing in these patients beyond just looking at the PSA levels. Again, just to summarize the data for our audience out there, that if PSA level is less than 0.5 in patients after surgery, and if they do not have access to any molecular classifiers, it is okay to not give them androgen deprivation therapy based on the POSEIDON analysis.
Jonathan Tward: That's how I would interpret it exactly, if they do not have access to further stratification. But the real value is we do have tools that can help further stratify.
Neeraj Agarwal: And if you have a tool, that was the next point, then we should be looking at these molecular classifiers and see if the luminal B type is present. If that luminal B phenotype is present, patients will likely respond to androgen signaling pathway inhibition.
Jonathan Tward: That's correct. For those patients with low PSAs that you'd otherwise now omit therapy for, we do think this is a subgroup of patients that greatly benefits. Within our own data that we presented here at GU ASCO, we actually showed that if we look at our luminal B patients at the Huntsman Cancer Institute, University of Utah and omit androgen deprivation therapy, their metastasis rate at three or four years is something like 40 or 50%. Whereas if we give them the appropriate androgen deprivation therapy in our hands, that drives it down to something like 10% at three years. It's a very, very potent effect that holds up for the androgen deprivation question component of it as well.
Neeraj Agarwal: With salvage radiation.
Jonathan Tward: With salvage radiation therapy. This is very particular to the post-prostatectomy setting. I do think that we need to be careful and not yet apply such a judgment to an upfront patient being treated with just radiation therapy for intact prostate. I think they're different contexts, but clearly in the post-prostatectomy salvage radiation space, this is really the first predictive marker that I think we can use to very clearly tell us that it's safe to either use or omit when PSA levels are under 0.5.
Neeraj Agarwal: That's great discussion, and thank you for sharing your insights on how to approach management of patients with biochemical recurrence after radical prostatectomy.
Jonathan Tward: Absolutely my pleasure.