Jeffrey Zhong: Yeah, thanks for having me here. It's really a pleasure to be here today.
Zachary Klaassen: And you've presented some great work at ASCO looking at the UNITE Study and really looking at the dose interruptions during EV/pembro. And so we'll get into that in a minute, but maybe just level set for our listeners why we're looking at EV/pembro. What was the data that sort of led to this first-line setting in metastatic urothelial carcinoma?
Jeffrey Zhong: Yeah, happy to go into that very briefly. So currently right now, EV/pembro is the current preferred first-line therapy regimen for advanced urothelial carcinoma. That's primarily based off of the results from the landmark EV-302 clinical trial where we showed essentially prolongation of progression-free survival and overall survivals in patients with advanced urothelial carcinoma who were previously untreated before.
Zachary Klaassen: That's great. The UNITE Study, we've seen some analyses come out of this. Just give our listeners what that cohort looks like and what the study design was for your analysis.
Jeffrey Zhong: Yeah, happy to go into that. So UNITE database or consortium is essentially a large collaboration among 17 US academic centers that was initially designed to help us try to better understand the real-world treatment experiences of EV and other potential target therapy regimens in patients with urothelial carcinoma. So at least with regards to our study that we performed, the main question we wanted to ask is if on treatment dose modifications of EV+P impact real-world clinical outcomes in patients with advanced urothelial carcinoma.
So to answer this question, we leveraged this UNITE database to conduct a retrospective analysis of patients with advanced urothelial carcinoma treated with EV+P. And for our study, we performed three primary analyses that I'll go into more detail in a little bit, but we excluded all patients who had an upfront EV dose reduction in our study. So we had a total of 360 patients in this retrospective analysis.
175 of these patients had an EV dose reduction. And out of those patients, 100 patients had a early dose reduction, which we defined as occurring less than nine weeks after starting EV+P treatment, and 75 patients had a late dose reduction, which we defined as occurring nine or more weeks after starting EV+P treatment. And then overall, 68 patients had a dose hold.
Zachary Klaassen: I see.
Jeffrey Zhong: Yeah. So kind of going into our analysis and what we found overall. So for our first analysis, we conducted a multivariable Cox analysis between patients who had an early dose reduction versus those without a dose reduction and also between patients who had a late dose reduction versus those without a dose reduction. And the facts that we adjusted for were age and then also prognostic factors that were chosen in accordance with the Belmont criteria.
And what we overall found was that there was a prolonged progression-free survival and overall survival in the patients who had an early dose reduction and late dose reduction versus those without a dose reduction. For progression-free survival and overall survival, we had hazard ratios in the 0.5s for the early dose reduction cohort and 0.3s for the late reduction cohort.
We think this was more suggestive of a possible immoral time bias that was going on because patients who have a dose reduction had to have by definition at least survived, most likely haven't progressed at least until the point of that dose reduction occurring, which can kind of superficially make it seem like they have a prolonged progression-free survival or overall survival.
So that's why we performed some additional analysis. So we did for our second analysis a time-dependent covariate and multivariable analysis between patients who had a dose reduction versus those without, and also between patients at a dose hold versus dose without. And this was also done to reduce that immoral time bias that I had mentioned previously about.
And we found in this overall, there's no significant difference in progression-free survival or overall survival between the patients who had a dose hold versus those without. However, on that similar analysis between patients who had a dose reduction versus those without a dose reduction, we once again saw a significant prolongation in the progression-free survival and overall survival.
Our initial hypothesis for why this is is number one, either our time-dependent cohort wasn't able to fully overcome that immortal time bias or what we think is more likely is that there is essentially another overlying possible selection bias that's at play, something like a healthy user effect or healthy user bias that could be screwing our results in favor of the dose reduction cohort. So that's how we performed our final and third analysis.
This was a complimentary multivariable landmark analysis that was aligned to the number of treatment cycles received with six analyses performed from cycles two to cycle seven. Once again, this was done to reduce that immoral time bias as well. And what we found in this analysis was that there was no significant difference in progression-free survival or overall survival observed between patients at a dose reduction versus those without at most of the landmarks, except for a dose reduction favorable OS at cycles four and five that had some borderline significance.
Zachary Klaassen: I see. That's a beautiful summary of a lot of analysis. So you came very well-prepared for our discussion. That's fantastic. Always when we look at these real-world analyses, these are patients that may or may not have been candidates for enrolling in these clinical trials and it's a very important topic in terms of whether there's dose holds or there's reductions, especially with that EV component. What would your message be to people that are going back to their clinic after ASCO in terms of interpreting your data for real-world application?
Jeffrey Zhong: Yeah, of course. Yeah. So I think at the end of the day, we also think about the patient in front of us and what we should do. And I think our study gives us some evidence that these dose reductions and dose holds may be a reasonable strategy for helping to manage some of these toxicities, but really decision to make these dose adjustments is really like a multifaceted decision that has to take account not only the toxicities, but also quality of life and also patient goals of care.
Zachary Klaassen: Sure.
Jeffrey Zhong: So I think our study really just helps to provide an extra data point with regards to having that kind of discussion with patients and ultimately making that decision to have those dose reductions or dose holds. I will say I do caution applying or generalizing our findings to all patients who are being treated with EV+P who are experiencing toxicities. So we did have some important limitations to this study. I think number one, it is a retrospective study, so we do not have things like randomization, no central radiology review.
It can also be susceptible to missing data that's not in the chart. And also, like I mentioned previously, we did run to some pretty significant immortal time biases and selection biases that we tried to address with the landmark analysis, but one drawback of that approach is that it further limits the power or generalizability of our findings as additionally exclude patients who aren't able to reach those specific landmarks or milestones.
Zachary Klaassen: And it's a fair assessment of the analysis. I think that's fantastic. Phenomenal presentation. Great job. Congratulations on the ASCO presentation and for joining us on UroToday. Really appreciate your time. You'll be applying to fellowship next year, so keep an eye out for your application for sure and thanks for joining us on UroToday.
Jeffrey Zhong: Yeah. Thanks so much for having me.