Andrew Armstrong: Thank you, Tanya. Thanks for interviewing me today about this topic.
Tanya Dorff: Yeah, so you're presenting more data from your excellent study, ARCHES, which was done in the metastatic hormone-sensitive prostate cancer space, looking at doublet therapy with enzalutamide. So in this particular analysis, you're looking at patients who stayed on treatment for a long time. Tell me a little bit more about why you undertook this, what you were hoping to learn.
Andrew Armstrong: Yeah, thank you. So ARCHES kind of transformed the hormone-sensitive space, led to the FDA approval of enzalutamide 2019. And then we updated the five-year data just this past year where we saw a three-year improvement in median survival. But some patients progress within two years. Some patients are remaining on therapy well past five years and we sought to identify pretreatment predictors of treatment duration. Now, treatment duration is a composite of efficacy. So you stop when you have radiographic progression, but it's also tolerability. So you might stop because of toxicity.
And so in a real-world sense, we wanted to look at what are the predictors of treatment duration. And so one of the most important things that I think that we found was that disease volume was the strongest predictor. However, high-volume patients were still very likely to experience long treatment durations, meaning five plus years. About 55, 56% of these patients were on treatment well past five years. So that's good news, that just because you have high-volume disease doesn't mean that you'll be progressing within two or three years. So that's good news for patients.
Tanya Dorff: That is encouraging. Was there anything different about those high-volume patients who stayed on long-term?
Andrew Armstrong: So in ARCHES, we don't have genomic data. We don't have data on the genetic makeup of tumors. And I think a lesson for industry is to really try to collect those tumors so that we can identify pretreatment predictors. We've looked at that in separate analyses like in our PROMISE registry that we do together. And we are starting to identify molecular fingerprints of resistance like PTEN loss, TP53, RB1 loss that can create that neuroendocrine small-cell transformation. That tends to be enriched in short-term responders, SPOP mutations conversely on the other side.
In ARCHES, we could really just focus on clinical predictors. And we didn't find any major predictors other than disease volume. Synchronous patients tended to do just as well. Age is certainly a factor related to toxicity. We saw that PSA nadir and objective responses, which is now looking at post-treatment outcomes, tends to be associated with those long-term responders.
So if you achieve an undetectable PSA and/or you get a complete remission on imaging, that's going to enrich for your probability of being on an ARPI five plus years later.
Tanya Dorff: So PSA nadir certainly has been a big story in this particular patient population. I'm interested about the imaging because I think there's still a lot of uncertainty in real-world practice about how often do we need to be imaging our patients once we start them on their doublet. So what were you doing in ARCHES and how would you apply that in practice?
Andrew Armstrong: So in ARCHES, we were scanning patients very regularly, even independent of their PSA or clinical state. We recently reported that discordant progression could happen in about a quarter of patients. So imaging progression without even a PSA rise. So the importance of imaging can't be overstated. Doesn't mean you have to do it every three months, but doing it somewhat regularly every six to 12 months, particularly in the first couple years, I think is important to pick up that discordant poor-prognostic event.
In terms of treatment duration, achieving that undetectable PSA was necessary but not sufficient, meaning some patients who even achieved an undetectable PSA could still progress within two to five years, but achieving radiographic remission at that six- to 12-month landmark, the two together kind of tell you that your patient's going to do very well.
Tanya Dorff: Well, that's all super helpful and practical information that I think our audience can apply. The other piece I just wanted to touch on briefly is you do mention that some patients do stop the ARPI due to toxicity. I feel like in my practice, patients are so nervous about even lowering a dose, let alone stopping, even if they're having quite substantial toxicity. So maybe you can tell us just a little more about that.
Andrew Armstrong: Certainly. In ARCHES, we did see an enrichment of toxicity in older patients and those with comorbidities, but these patients still do very well long-term. And I think also at ASCO, we're going to see some comparative data across different ARPIs for different tolerability. But I would emphasize that dose reductions can help you get around this.
With enzalutamide, if you have fatigue or neurocognitive effects or effects on frailty falls, dose reduction, dose holds can really help you get around that. Exercise, attention to bone health to reduce fracture risk, really cardiovascular fitness is some of the major drivers of treatment discontinuation, particularly in those first two years.
Tanya Dorff: Yeah. I think that's really important to emphasize what the patients can do proactively, empower them.
Andrew Armstrong: Absolutely.
Tanya Dorff: But also, yeah, that if someone's really hitting a wall, taking a little break, then dose reducing and continuing would be my preference for sure.
Andrew Armstrong: Exactly.
Tanya Dorff: Yeah.
Andrew Armstrong: Yeah.
Tanya Dorff: Well, thank you so much for sharing these results with us.
Andrew Armstrong: Thank you. You're welcome.