Alicia Morgans: Hi, this is Alicia Morgans, GU medical oncologist and Associate Professor of Medicine at Northwestern University in Chicago, Illinois. I am so excited to have here with me today, Dr. Neal Shore, who is the Director of the Carolina Urologic Research Center in Myrtle Beach. Thank you so much for being here.

Neal Shore: Thank you for having me, Alicia.

Alicia Morgans: Wonderful. So I wanted to speak with you about the ASCO presentation on the ARAMIS data, which is really an update on the Phase III trial that we've heard about in the past looking at darolutamide in a non-metastatic castration-resistant prostate cancer population. And this was the overall survival data that has finally become statistically significant. Can you tell us a little bit, just a recap, of what the Phase III trial looked like looking at darolutamide?

Neal Shore: Yeah, so the ARAMIS trial design looking at a primary endpoint of MFS, and then eventually a final analysis on OS was taking patients who had non-metastatic CRPC. So a rising PSA in the setting of a castrate level of testosterone and by conventional imaging, CT scan and technetium bone scan, no evidence of a positive disease with the exception of potentially a lymph node in the pelvis below the bifurcation they could be allowed in. And they had to have a PSA doubling time of less than or equal to 10 months.

As you stated, this was a trial that met its primary endpoint of MFS and the paper was published in the New England Journal and also presented at ASCO by Karim Fizazi. And it was a two to one randomization of a little over 1,500 patients receiving darolutamide 600 milligrams orally, twice daily with ADT versus a placebo-controlled combination with ADT as well. In ARAMIS, the MFS endpoint was met with a hazard ratio of 0.41. And so I think that endpoint, just to remind our viewers, there's a composite endpoint of overall survival as well as radiographic progression.

But when we did that primary MFS analysis, very similar to the SPARTAN and PROSPER studies, there were very few or only around 20% or so of the events that were from death or mortality. And so the OS data at the time of the New England publications as well as the time of FDA approval was still immature. So what Karim Fizazi presented at ASCO Virtual this year, on behalf of all the authors and the investigators, was our final analysis on the OS data. And very nicely the hazard ratio is 0.69 with a p-value of 0.003, demonstrating the survival benefit for patients in the darolutamide arm.

Alicia Morgans: So thank you for sharing that with us. I think one of the reasons that I find this so compelling and such a powerful endpoint is that this nonmetastatic CRPC population inevitably in the control arm went on to further life-prolonging therapies in most cases. Most likely given the stage of disease that was being studied and the survival benefits still held up, which I think is really important and is helpful to us as we counsel patients. It's not just that we're preventing and prolonging the time to develop metastasis. We're actually helping you live longer, regardless of what we end up doing after this initial treatment. Is it something that I'm sure that you find it a compelling endpoint as well?

Neal Shore: Oh, sure. Yeah. I'm so glad you brought that up. Yeah. I believe the number is more than 60% of patients crossed over in the control arm and were able to get darolutamide or other life-prolonging approved advanced prostate cancer agents. That's really important because that obviously affects the delta of the survival. But even with that, it's clear that the treatment arm, the darolutamide effect starting from baseline met an overall survival benefit.

Alicia Morgans: Yeah. Very, very important. Some other endpoints that were reported at Virtual ASCO 2020 that we should definitely mention are things that affect the quality of life of patients and are certainly important to their day-to-day. Things like time to symptomatic skeletal event, time to next cytotoxic chemotherapy, and the time to pain progression. Can you comment on those?

Neal Shore: Yeah. Well, all three of those that you've nicely summarized met statistical significance. As it relates to the time to pain progression, and we recognize that these patients who are nmCRPC are for the most part asymptomatic, they don't have tumor burden. They may have some side effects from their ADT. And when we looked at the Kaplan-Meier in this analysis, the hazard ratio was 0.65 and from time to pain progression from a baseline in favor of darolutamide. So this is something that's always very important to patients. We saw the MFS difference between treatment and control arm of nearly two years. Similarly, we saw approximately about a 15-month advantage to avoidance of pain progression in the treatment arm. So this is something that many of us don't focus on as much as patients will focus on. They want to live well. They want to avoid pain and they want to live a high quality of life and pain is always a great concern.

Similarly, we looked at the time to the requirement for cytotoxic chemotherapy and in the darolutamide arm versus the control placebo arm, the hazard ratio was 0.58. Again a global study, and so there may be some regional differences and some practitioner differences, but be that as it may, cytotoxic therapy, which certainly has its role and its appropriate advantages, is something that when it can be avoided is beneficial and certainly looked forward to by patients. So the hazard ratio of 0.58, I thought was very compelling.

Alicia Morgans: Great. The other thing that I always think about when I'm thinking about patients' quality of life are the adverse event profiles and you and a team from Duke and others worked together to do a really interesting approach to a method to attempt to do a cross-trial comparison of adverse events. You used this matching adjusted indirect comparison method to look at the darolutamide side effect profile against apalutamide. Can you tell us a little bit about that?

Neal Shore: Yeah, no. It's a bit of a mouthful, isn't it to say that? Just to also further add when we looked at the final analysis in the adverse event profile when we achieved the OS benefit, we didn't see any new, additional adverse events or safety effects. The main things that we saw in the ARAMIS trial were issues of a slight difference in fatigue, but overall we found essentially no differences of significance in neurocognitive effects. That's one of the reasons I got particularly interested in the darolutamide molecule because of its different molecular structure and lower risk of crossing the blood-brain barrier in preclinical models. This also explains why in our early phase trials and in the Phase III, in comparison to the other drugs in the space, it always allowed in patients who had a history of seizure or even epileptogenic or seizure risking medications.

So what we wanted to do in this poster that we presented at ASCO this year, the first author is Shan Jiang and the title is "Safety outcomes of darolutamide versus apalutamide and enzalutamide in nmCRPC" and we use this methodology known as matching adjusted indirect comparisons. And it's really a way to say, "Okay, we haven't yet done a direct prospective head-to-head comparative study, but what can we glean from the literature or the published literature? How can we look at the information that was described in publications and how can we sort of also look at the verbiage of the data dictionaries and how the data was collected?" And it's a bit complicated. I would really instruct listeners if there's a link they can go to, to review the details of this methodology. But it's pretty rigorous in carving out an adverse event or a descriptor when it's not consistent across two trials in comparison or three trials in comparison.

So ultimately we looked at the three New England publications. We looked at differences in safety outcomes of daro versus apa, and daro versus enza. And I can tell you that there are, as I already mentioned, the strengths of this is to try to understand across trial differences without doing a prospective comparative study. We did allow for what I described as multiplicity adjustments according to statistical procedures. So I think that those caveats are important and as well as the strengths, but some of the limitations are that the only known baseline factors that were consistently reported were, they were only included as covariates in this MAIC type of methodology.

And again, one has to recognize these are trial studies as opposed to real world. But the punchline, or at the end of the day, after doing a lot of crunching of the data and trying to get an understanding so clinicians can say, "Well, I've got these three drugs. How do I decide which one to use? And how can I inform my patients until I have the advantage of a prospective comparator study?" After we did our adjustments, what we concluded was that darolutamide showed a generally favorable safety profile in comparison to apa and enza. It had a statistically lower risk of falls, fracture compared to apa and a lower risk of falls, dizziness, mental impairment, fatigue all grades, and severe fatigue compared to enza. And this goes back to my comment regarding, or at least why this could be is because of a potential lack of blood-brain barrier penetration.

Alicia Morgans: So I think that's a really interesting methodology and an interesting way to compare, though it's certainly hard to compare across trials. This is one approach to try to do that in a thoughtful way that minimizes the confounding that happens because we are doing completely separate trials. This isn't a prospective randomization of one drug versus another, but really thought-provoking information. And I think it's, like you said, useful in conversations as we have them with our patients. So as you think about the ARAMIS data about your other work, in terms of side effects and safety profiles, do you have any closing thoughts for the listeners?

Neal Shore: Yeah. I think that the world of how we approach GU oncology patients, I think for all the patients with advanced malignancies, now we have to consider this notion of neurocognitive effects. Whereas heretofore, I think we were kind of a bit uncertain how to approach it. I think we were a bit squeamish on how to figure out what are the best ways of evaluating for neurocognitive impairment? What are some of the better instruments that we could be looking at? How do we better design studies? How do we look at the reported adverse events and safety tolerability information? There is potentially some imaging work that can be done. I'm very excited about the whole concept of neuro-oncology and how we can all work together, just like we were doing in cardio-oncology, and think how do we really optimize patient care?

Alicia Morgans: Wonderful. Well I, again, appreciate your efforts in helping us make smart decisions with our patients and for talking with me a little bit about this survival data for darolutamide. I think it's really important and a nice way to have that further information for patients, that it's not just metastasis-free survival, it's actually overall survival, which makes a huge difference. So thank you so much for your time and your expertise today, Dr. Shore.

Neal Shore: Thank you, Alicia.