The Use of Biomarkers in the Adjuvant Setting to Understand the Effect of Nivolumab in the Checkmate 274 Study - Andrea Necchi
March 16, 2023
Andrea Necchi, MD, Medical Oncologist, Department of Medical Oncology, Fondazione IRCCS - Istituto Nazionale dei Tumori, Milan, Italy
Alicia Morgans, MD, MPH, Genitourinary Medical Oncologist, Medical Director of Survivorship Program at Dana-Farber Cancer Institute, Boston, Massachusetts
Alicia Morgans: Hi. I'm so excited to be at ESMO 2022 with Professor Andrea Necchi talking about CheckMate 274 and the presentation that he gave. Thank you so much for talking with me today, professor.
Andrea Necchi: Thank you, Alicia. Thank you for inviting me.
Alicia Morgans: Of course. So you were able to, in your presentation, really share some of the investigation that you and the team did into understanding how we can use biomarkers in the adjuvant setting to understand the effect of nivolumab per this study. So tell us a little bit about the study and its initial findings, and then what you and the team investigated in this biomarker work.
Andrea Necchi: Thank you, Alicia. So I was really proud to present the biomarker data translational part of this practice changing study, the Checkmate 274 study investigated that the adjuvant nivolumab therapy compared to placebo in patients with high-risk urothelial cancer after radical resection, radical cystectomy, radical nephrectomy, providing a statistically significantly and clinically meaningful advantage in disease for survival, both in all comer population and the PD-1 positive population and leading to the registrations of nivolumab, the approval of nivolumab by the FDA for all comers and by EMA for PD-1 positive patients. So the translational biomarker part at the median follow up of 23 months in this study provided evidence that there may be additional biomarkers, additional players in the game of identifying patients who are more suited for adjuvant immunotherapy approach compared to those who may be judged as cured by surgery alone. So in this way, we investigated, thanks to the older team, we investigated a lot of potential additional biomarkers through different ways of analyzing tumor tissue, all tumor tissue based on the radical resection before starting therapy.
So biomarkers were analyzed at the gene expression [inaudible 1:59] with whole exome sequencing with the calculation of the tumor mutational burden and bioimmunohistochemistry. So several signature scores were analyzed. Scores and signatures are related to the immunotherapy effect to preexisting immunotherapy like the interferon-gamma gene expression signature that was originally developed in Checkmate 275. So with nivolumab being patient with more advanced disease, providing data on the association with survival benefit with nivolumab and expression of single genes like CD4 and CD8 that are linked to antitumor immunity and to T cells that are actually the effectors for immune system to act against the tumor cells. In summary, the data were in favor of identifying an association between interferon gamma signature with the both prognosis and the treatment effect. So as a predictive factor for nivolumab efficacy in the adjuvant setting. The same for CD4 gene expression. It was found to be both prognostic and potentially predictive of nivolumab efficacy because higher level of CD4 gene expression, as well as higher level of interferon gamma signature score, were associated with improved disease-free survival with nivolumab arm and not with placebo arm.
So providing that of predictive efficacy and predictive effect, not just a prognostic effect. For CD8 expression, it was evaluated here as with digitally immunohistochemistry rather than gene expression. The data were more in favor, prognostic effect, meaningful difference between arms there was not identified, so a predictive value here was weaker compared to the other biomarkers and it is more or less the same with the tumor mutational burden. So a higher level of multi-tumor mutational burden were associated with the effect, with a trend towards a higher effect of nivolumab compared to placebo.
But the difference was not striking, was not statistically significant, but there was a trend. So the issue here is to try, as a next step of course, it's just, of course, the tip of the iceberg of the enormous amount of other signatures that have been developed and investigated within the biomarker part, of course, and that will be expanded in the manuscript, that will follow the presentation. But the key next step, the key issue here is to try to identify how to combine these biomarkers and to identify a way to develop a composite biomarker approach and composite a tool in order to integrate all of this evidence towards the patient selection basically.
Alicia Morgans: Absolutely, and I think it's so interesting and important just to reiterate that these markers were all assessed on pretreatment biopsy specimen and despite that they were able to serve as both prognostic and predictive markers here. So really important that they could show a cohort that's destined to do better no matter what. That's the prognostic biomarker. But then also to show, predict which patients may benefit even more from nivolumab. Now, did you find in this initial study, in this trial, that there were patients who did not seem to benefit from adjuvant nivolumab based on this work?
Andrea Necchi: Yeah, it's an important question. So I'm looking at the things the opposite perspective of patients who are non-responding or not getting benefit with nivolumab therapy. Of course, among the amount of signature that are being developed, we looked also the signature that they are found from previous studies in other disease settings to be associated with lesser benefit of immunotherapy like the TGF beta signature score or other similar signatures. But in fact, in this case, we did not find it is striking association between both prognosis or prediction of efficacy or inefficacy of the drug. So the things here seem to be a bit more complicated compared to the metastatic setting, of course, and it's likely that we have to move outside of the gene expressions course towards, as I said, developing composite models or looking at totally different things like ctDNA or other biomarkers that will be presented that will be the objective of further analysis from this large study.
So it's an ongoing process, of course. The main point here to make is how to best rescue the proportion of patients with a PD-1 negative tumor who seem to benefit lesser from the nivolumab therapy compared to the PD-1 positive tumor, despite that the drug is still approved for all comers in the United States. Of course that the average ratio 0.5 for PD-1 positive patient for disease for survival is quite more compelling compared to the PD-1 negative population. So the distribution, an important finding from that I presented from this study, is that the distribution of the biomarker expression in signatures for each one of the biomarkers that presented between the population of patients with negative or positive tumor was pretty much similar. Meaning that by adding additional biomarkers towards PD-1 expression by immunohistochemistry, we may be able to rescue the population of patient with PD-1 negative tumor to benefit the most from nivolumab therapy. But it is just an assumption. Of course, we are still tickling a small part of the entire body of what can be identified as potentially useful for the translation in the real world.
Alicia Morgans: Absolutely. So I think it's really important and interesting, just to emphasize, that certainly we are finding signals where we might find patients benefiting, more ultimately in an adjuvant trial. We certainly also want to know who can be spared this treatment if there's not going to be benefit and work is ongoing. I really love and appreciate that you and the team are working also, not to look at single biomarkers, but to look at signatures and try to understand these in compilation over time as well and look forward to you looking at these signatures and single biomarkers in other studies, in other populations, and also pulling apart the adjuvant versus metastatic populations and understanding the differences there. I think it's really incredible as we expect that some of these biomarkers at least, may be things that are going to be changeable over the course of disease and may be something that could be up or down regulated. And so wondering how they change over time and that may be important will also be interesting. Where do you see the field going? Where do you see this work going next?
Andrea Necchi: The main point here is to try to put our efforts in the way to contextualize the findings into the overall developments of biomarker for patient selection in bladder cancer because we have contrasting signals across the different stages. For example, for interferon gamma gene signature, we did not have any strong signal of association with survival benefit for the NABUCCO study, that was a neoadjuvant study, not an adjuvant study with the nivolumab ipi/nivo before radical cystectomy. But at the same time we had positive results for association in metastatic stage from Checkmate 275 and then positive association in the adjuvant setting for 274. So how to best interpret this data with, it is still basically unknown. And this partly dependent on the changes of the tumor that are induced by therapy. So for example, when looking at the tissue post neoadjuvant chemotherapy, as most of the patient have received in the Checkmate 274, had received some neoadjuvant chemotherapy before the radical surgery, things may have changed from the tissue from TURBT, so the tissue from the pre chemotherapy setting.
So the integration of the data from the overall studies that are testing biomarkers in bladder cancer throughout the clinical stages will be quite tough and it's, of course, a next step and the final step, perhaps maybe that the effort of integrating the tumor biomarkers and liquid biomarkers in the effort of identifying the best candidates for perioperative therapy, adjuvant or neoadjuvant therapy, and of course integrating [11:25 inaudible] biomarkers like those investigating in this study with, for example, the evidence from the IMVigor study with ctDNA effect on the association with survival would potentially, theoretically lead to the best results possible for the patient selection in this case. Of course, the issue of providing patient potentially toxic therapy that may be useless for a small proportion of patients like patients who are cured with the radical surgery alone is a very important issue of everyday practice and accumulating data that may help us in guiding the discussion with the patient and the patient decision making would be a key for moving from clinical trial data to the real world practice.
Alicia Morgans: Absolutely, and I think ultimately that's where we need all of this work to go to really help us get to the real world practice and how we care for these patients. So I appreciate the work that you're doing and thank you so much for sharing your insights on this presentation for ESMO 2022 and your thoughts on really where we need to go in the future. I always appreciate your time and your expertise.
Andrea Necchi: You're welcome. Thank you.