Sophie Merrick: Hi, thank you for having me.
Pedro Barata: Absolutely. Thank you so much for joining us. And the idea today is to talk a little bit about your very well done presentation on patient-reported outcomes for RAMPART. So just for folks who are less aware, RAMPART is another adjuvant effort, is an investigative initiative study run in England, and looked at two different IO-based approaches, durvalumab and tremelimumab. There's actually three arms for this study. A few months ago, we've seen data for the combination of the durvalumab tremelimumab. So that's a PD-L1/CTLA-4 combo compared to observation in that patient population. And Sophie, you did a good job, a very good job I would say, with the study in regards to looking at how these patients do while on therapy. So maybe I guess we'll start there. Can you tell me why patient-reported outcome data you felt was relevant in the context of receiving immunotherapy without evidence of disease? In other words, role of weight of PRO in the adjuvant setting, and then how you apply that to RAMPART, and maybe we'll go from there.
Sophie Merrick: So I think patient-reported outcomes are important in both settings, but I think particularly in the adjuvant setting where you've got patients who may be cured of their disease anyway, and then you're thinking about adding a treatment on top with potential toxicity, then understanding the patient experience is really important. And obviously we have clinician-reported data, but nobody can tell us about the patient experience without hearing that directly from the patient themselves. So the patient-reported outcomes are really complementary to the efficacy results. And I think incredibly important if we're going to be discussing this treatment option and the impact of that with patients and as I said, particularly in the adjuvant setting where the risk-benefit is perhaps a little different.
Pedro Barata: Yeah, no, that's fantastic. I agree with you. And I think the threshold for patients to tolerate interventions is different than when you're facing with stage-four disease that you know that you're seeing tumor burden, that by itself will cause symptoms. So I couldn't agree with you more. So tell us really, what are the findings of this evaluation?
Sophie Merrick: So we measured patient-reported outcomes at baseline and at week 16, which is probably when we can anticipate some of the higher level of toxicity is going to be, and then at month 15. And just to make clear, the treatment is for one year. So that was three months after they completed therapy. And when we looked at the change from baseline to that first earlier time point at week 16, we saw a kind of general worsening in functioning and symptoms. And the ones that were statistically significant and clinically meaningful were overall health and quality of life, role functioning, and fatigue and insomnia. And then when we followed that through, when we looked at month 15, so after completion of therapy, a lot of those areas appeared to improve, but we did see later emerging effects in cognitive function and in pain as well, and those were both statistically significant and clinically meaningful. And it's just worth noting with the cognitive function, that's self-reported on the questionnaire. That's not a subjectively testing cognitive function. So there could be broader reasons for that kind of change in cognitive function that we're seeing there.
Pedro Barata: Right. So super important findings in my opinion. And this gets a little bit into the weeds of, my question gets into a little bit of the weeds of how do we define no difference between the scores versus what is considering worse? Because I think most of us do struggle to be more evaluating PRO data overall. And sometimes the cutoffs are different or the difference or the delta is different in terms of the scores of whatever parameter we are evaluating. Can you just walk us a few, just for context, what was considering a delta, what was the cutoff for us to say, you know what, this is impacting patients in a negative manner that you've seen this here in RAMPART.?What was the cutoff used and whether or not that's the same cutoff being used in other studies? Why that cutoff and not a different one?
Sophie Merrick: Yeah. So we use clinically meaningful difference thresholds, which are kind of established in the EORTC literature. But historically a lot of trials have used a kind of blunt 10-point threshold cutoff for what is a clinically meaningful difference. And that is what you will see in a lot of trials. And so it's difficult when you're doing these kind of cross-trial comparisons anyway, but even more so with the PRO data when different thresholds would be used. But there's increasing evidence that thresholds are different depending on the different domain, and that you can also categorize those into small, medium, or large impacts that are clinically meaningful. And so those are the thresholds that we apply to this analysis, the Cox thresholds. And so they provide a little bit more granularity to the data and also differentiate those different domains and different impacts that we might be seeing. And if we had used that 10-point threshold, which historically was what a lot of trials had used, we wouldn't have seen any of these impacts as clinically meaningful. So it was only by applying these kind of more up-to-date thresholds that we did detect those changes, which I think is consistent with the pattern of toxicity. So I think we did show probably what was impactful for patients.
Pedro Barata: And so we're clear for folks who are hearing this, basically it's a five to 10, small clinically meaningful change, and then 10 to 20, moderate change, and then more than 20, large change. Is that correct for the EORTC or is it a different threshold?
Sophie Merrick: So it actually varies by specific domain. So it will be different for overall health and quality of life, for role functioning, and for each symptom as well as to small, medium, and large.
Pedro Barata: Right. So depending on the threshold that we'll use, there is a possibility that how we call out this is detrimental versus it may not be as detrimental and we might even read it out as no difference is almost, I'm not going to say it's pre-specified, but obviously depending on the threshold we choose, we might get a different outcome. Is that a fair?
Sophie Merrick: Exactly. Exactly. And so we predefined all of these in our statistical analysis plan per domain and with these different levels of thresholds. But yes, if we had used a 10-point cutoff, then we wouldn't have seen any of the changes.
Pedro Barata: Right. So to me, that's actually a super, super relevant point because I think the world and our community is adapting to showing PRO and interpret that in the context of efficacy. And I'll get back to that point. And we see over and over again, studies that do use a lot the 10-point threshold, and that's why I was getting there. I appreciate the thoughtfulness of your response. And to say it's really relevant to go beyond what everybody considers that delta to understand. And going back to your presentation, you then have the PRO data and you look at efficacy right now at the DFS perspective and you really call out the balance, but we have to make, okay, should we consider this intervention, which to me is very positive in the high-risk disease, although it was defined in a different way from what we have learned how to call out high-risk disease based on the KEYNOTE-564 perhaps, so the definition of high-risk is different there, but to me it's quite positive, but we don't have OS data yet. So my question to you is, yes, we have to use these data in the context of the DFS benefit to think of whether or not these will be available to patients. My comment, my question to you is, do you think that balance might change depending on how overall survival will read out on the same arm, durva treme, or do you think we might not even see that and really we need to focus on preventing cancer from coming back versus the toxicity of a dual IO regimen? What are your thoughts on that?
Sophie Merrick: So I think we certainly, it will be important to see OS data when we have it. I think we also will need to have a look at the comparison of durvalumab monotherapy versus active surveillance because that's going to give us an idea as to the contribution of components here and what tremelimumab is contributing. So I think that will all contribute towards the picture. I think the PROs are still important. I think if durvalumab and tremelimumab were to be approved, which it isn't yet, I think that this is going to be important in terms of conversations to have with patients particularly around the early toxicity that they can expect and also perhaps later emerging effects that we need to monitor and thinking more about survivorship care for patients as well, which perhaps we need to work on more now we're treating with immune checkpoint inhibitors in this early disease stage setting as well.
Pedro Barata: Gotcha. So before I let you go, I really want, can you give us a sneak peek, I guess, or a teaser regarding what we should expect or when should we expect data for durvalumab monotherapy? Can you give us an idea of what's coming for RAMPART?
Sophie Merrick: Yes. So the durvalumab versus active monitoring comparison, we should see results later this year. I can't give any more details on that. And the patient-reported outcomes should be with that later this year as well.
Pedro Barata: Kudos to you, Sophie. It's really relevant to see patient-reported outcomes in these contexts, as you said. Thank you for explaining to us and to me specifically regarding the importance of thresholds and what that meant to the patients who enrolled in RAMPART. And yeah, we're looking forward to see the next steps with the durvalumab, as you call it, CoC, contribution of component piece super relevant, PD-L1, which I'm sure all of us are going to try and put it into context of what we know so far and what's coming for the future from that perspective. So thank you so much for taking the time.
Sophie Merrick: Thank you. Thank you.
Pedro Barata: Thanks. Have a good day.