The First Neoadjuvant Trial Using Combination Immune Checkpoint Inhibitor and VEGF Inhibitor in Patients With Locally Advanced High-Risk RCC (NeoAvAx) – Axel Bex

May 24, 2022

Axel Bex joins Alicia Morgans to discuss the results of the NeoAvAx trial examining neoadjuvant avelumab and axitinib following nephrectomy for patients who are at high risk for recurrence. The NeoAvAx trial tests the neoadjuvant use of a combination of the immune checkpoint inhibitor avelumab and the VEGF-inhibitor axitinib.

Biographies:

Axel Bex, MD, PhD, Urologic Surgeon, Royal Free London NHS Foundation Trust and UCL Division of Surgery and Interventional Science

Alicia Morgans, MD, MPH, Genitourinary Medical Oncologist, Medical Director of Survivorship Program at Dana-Farber Cancer Institute, Boston, Massachusetts


Read the Full Video Transcript

Alicia Morgans: Hi, my name is Alicia Morgans, and I'm a GU Medical Oncologist at the Dana-Farber Cancer Institute in Boston. I'm so excited to have here with me today Professor Axel Bex, who is a Professor of Urology at the University College of London, as well as being a Urologist at the Netherlands Cancer Institute. Thank you so much for being here with me today.

Axel Bex: Thank you.

Alicia Morgans: So I wanted to talk with you a little bit about your presentation at GU ASCO 2022. Can you share with us what you studied and what you learned?

Axel Bex: So we presented data from a single-arm phase II study of neoadjuvant avelumab and axitinib in patients with high risk of recurrence in clear cell renal cell carcinoma. You may know that this is still an unmet clinical need. In this patient population, they recur frequently. Neoadjuvant and adjuvant strategies need to be explored. Now in the past, we had axitinib monotherapy single-arm studies, which showed a reduction in the tumor size in up to 43% of the patients. But unfortunately, these concepts were not taken further into randomization. I think it's also consistently difficult to do that, so it's easier to do adjuvant trials. But now with these new strategies coming up, immune checkpoint inhibitor combination therapies, we would like to at least lay a foundation to get some signals that would inform us about how we could proceed into randomization with these studies. Therefore, I think single-arm studies are very ideal.

So we looked at the primary endpoint of this study, which was primary tumor response. So we wanted to show that the RECIST 1.1 partial response rate was at least 25% of the patients included. We chose a classical Simon's two-stage design with 40 patients. And secondary endpoints were disease-free survival, safety, and toxicity. In addition, we did exploratory biomarker work, which is currently ongoing, which hopefully would inform us about what is the best patient group to do that. Now the new thing about this study was actually that you can not use postoperative risk scores like [inaudible] because that's a neoadjuvant thing. If you pretreat them, they change their size and therefore might have differences in these risk scores. So we used a hybrid system of clinical TNM staging and [inaudible], and only the higher stages or if they were lower stages, but the higher firm and grades would be included for a truly high-risk patient group.

That is also shown if you look at the inclusion of the patients. 90% had at least a T3A and higher and 42.5% had clinical lymph node-positive disease with a mean diameter of 2.6 centimeters of these lymph nodes. Now patients would then be treated for at least six weeks with continuous axitinib, which could be titrated up to 10 milligrams twice daily. And they would receive three courses of avelumab, 10 mg per kg, intravenously every two weeks. Then they had an interval staging, and if they would progress, they would be taken off of the medication and receive surgery. Otherwise, they would continue for at least three months. Now, why three months? Because retrospective data and later also exploratory data from the prospective metastatic study showed that it would take about four and a half months before these big tumors would actually downsize properly.

On the other hand, you need to keep a balance because you do not want to harm patients and take them off for surgery too long. At the end of this period, surgery would then be performed 36 hours after the last dose of axitinib. The primary endpoint was met. So we had about 30% of the patients who had been downsized by RECIST as a partial response to their primaries. And they could be quite large. They were ranging from 7.8 to 18.8 centimeters with a mean tumor diameter of 10.3. So the median downsizing of these tumors was 20% and none of the primary tumors progressed. In terms of disease-free survival, we have a median follow-up period of 23.5 months. 13 patients recurred, which is 32.5% of the population, and which is a little bit what you would expect in this situation.  So whether they actually have a benefit or not, we are unable to tell that without randomization.

Safety is very important in these studies. So we looked at delays of surgery. I'm happy to say that there was only one patient who had a delay of three weeks because of a hyperthyroidism grade two. There was one patient who seemingly progressed with his primary tumor, but which then turned out to be a biopsy-induced hematoma. And there was one patient who unfortunately progressed to liver metastasis during treatment. In terms of surgery, surgery time, estimated blood loss, as we would expect in this locally advanced disease stage, and 25% of the patients had severe desmoplastic reactions. This is not as much as has been reported from retrospective studies. That said, for example, in France, after Ipi/Nivo, they reported would happen in 80%. This wasn't our impression.

13 series of adverse events occurred in this study. They were attributable to prolonged hospitalization and readmission and only a minority to actually trial procedures. And if you look at the common overall toxicity, which was treatment-related, it is exactly as we would expect from the JAVELIN data from the trial in the metastatic setting. So no grade four to five AEs occurred, and we had the impression that was a slightly higher infusion-related toxicity rate, which was 27.5% for grade one and grade two.

Regarding biomarkers, it was a bit disappointing. We found, and we could also reproduce, that which we had published earlier or presented earlier at ASCO last year, that in the unselected population, there was an increase of PDL1 and CD8 densities. And when we compared the post-treatment tissue to the pre-treatment biopsy, now this is something that others have reported as well, but what we couldn't show is a significant difference between those who had a partial response in their primary tumor and those who hadn't. What we, however, could show is that there was a significantly reduced CD8 density in those who recurred. This is quite interesting, this has also been shown by others. What gives us now an opportunity to investigate this further with spatial transcriptomics and with a few patients we already have investigated, we show that the ones who have a recurrence, interestingly show areas, intra-tumor focal areas, where there is a difference in immune responsiveness. So some patients have in that tumor, a high immune signature and very close to it, areas where there is a lack of immune signature, which is in keeping with almost an immune desert.

Alicia Morgans: So I think that is really fascinating. And as you said, at certain parts of that discussion, it can be challenging to think through how much of a benefit these patients have had in this single-arm study. How do you plan to investigate this further if you do, to really kind of get a sense of whether this is a reasonable approach for patients?

Axel Bex: Yeah. I think that's a very important question. There are two issues in my point of view. First of all, we would like to understand the mechanisms of resistance, so in those who it didn't work. I think more importantly for us clinicians would be, can we identify those patients who would have a benefit of this treatment actually before we start treatment? So most of the markers we identified are actually on treatment changes, but that's when you are already on the goal. So what you would like to know is who is the ideal candidate. And I think before we take this into randomization, we would at least like to understand who would be the ideal candidates.

There has been an interesting paper by the Crick Institute on an ADAPTeR study, as it was called. They looked at 15 patients who were primary metastatic and had been treated with nivolumab monotherapy in their primary tumor places. They took multi-regional sampling either before the start of the treatment and then afterward, and they could actually show that in responders, there would be a preexisting T-cell clone, which would then be expanded. And in those who wouldn't respond, this preexisting T-cell clone would not be responsive to neoantigens. Now, this is very difficult to prove. You need to do a tumor biopsy, and this is not in the clinical setting yet. So I think other possibilities would be to look at peripheral blood mononuclear sites and then investigate whether we find the same patterns there, but this is the future.

Alicia Morgans: Absolutely. And so as we sort of wrap this up and you think towards the future, but also want to just sort of wrap up what we discussed today, what would your message be to clinicians who are still curious and hopeful for a neoadjuvant approach to downstaging these really large kidney cell cancers?

Axel Bex: I think there are two opportunities here, for a urologist, they are usually interested in downsizing tumors, not so much downstaging, but there are some imperative indications, like patients who have single kidneys and have central tumors and you might avoid dialysis. But I think most importantly, is this outcome-related effect, and currently, we have to say, without randomization, we probably cannot say anything about it. I would definitely avoid treating these patients outside clinical trials, but I think the DFS signals that we saw, and I haven't told that yet, but I presented during ASCO, is that in patients who had a partial response in their primary, only two patients progressed or recurred. And those who hadn't had a partial response, the other 11 recurrences were noted. Now, this is not a statistically significant difference at the moment, but there is a trend and this patient population is small. So I think that if we are able to identify these patients who would have a benefit, then we would be ready for randomization.

Alicia Morgans: Well, I am really looking forward to hearing where this goes and to hearing the next steps that you and the team take to, as you said, identify the patients before they start treatment and understand which patients might benefit most. And then, of course, to see the results of this potential randomized trial that is really going to investigate this further. I sincerely appreciate your expertise and your time today.

Axel Bex: Thank you very much.