This transcript was automatically generated from the video recording and may contain inaccuracies, including errors or typographical mistakes.

Expert Guidance on Navigating the Evolving Head and Neck Cancer Treatment Continuum: What Matters Now and How to Apply It in Practice

I will try to set the stage prior to hearing some of the more recent advances in the head and neck cancer, continuing by discussing biology, prognosis and unmet need.

Case: 49 yo W with recurrent HNSCC

This is a patient of mine, a 49-year-old lady with a recurrent head and neck cancer. This was a woman who initially was diagnosed with a very advanced floor of mouth cancer, and underwent a very large and quite morbid surgery initially, with very poor pathologic features, pT4N3b, six positive nodes, positive ENE.

This patient received adjuvant chemoradiation with high-dose cisplatin. Despite that, three months later, unfortunately already developed new left neck adenopathy, which was confirmed recurrent disease. This patient received locally three cycles of chemoimmunotherapy with carboplatin/5-FU/pembrolizumab, and unfortunately developed primary progressive disease.

When I met this patient as she was looking for clinical trial options, you can see clinically and radiographically what we were facing. I will come back to this at the end of the introductory comments.

HNSCC Epidemiology

Head and neck cancer epidemiology. About 70,000 cases annually in the US. We know that this is a disease of high value real estate and complex anatomy of the head and neck, essential for speech and swallowing and function. As a result, is associated with substantial symptom burden. There is substantial toxicities associated with multimodality treatment in locoregionally advanced setting, which is the setting where we treat most of our patients.

Yet despite that, we know many of these patients develop recurrent or metastatic disease, more of an issue in HPV-unrelated biology as compared with their HPV-positive patients for whom we actually managed to cure the vast majority of these patients.

HPV-Associated vs Unrelated Represent Distinct Biological Entities

These distinct biologies are indeed distinct in terms of their biological as well as their clinical and phenotypic features. HPV-associated is driven by human papillomavirus and causes oncogenic transformation through the viral oncoproteins E6 and E7, which downregulate p53 and RB, the tumor suppressors, which leads to oncogenic transformation. As a result, this is characterized by, in many patients, the lack of carcinogen exposure, although some of these patients can also have tobacco exposure as well, is rising in incidence and, as we know, has a very improved survival.

HPV-unrelated or HPV-negative head and neck cancers are of course most commonly associated with carcinogen exposure, although we do see these now in patients without classic carcinogen risk factors and has a worse survival in locoregionally advanced head and neck cancer, as well as in the recurrent metastatic setting.

Key Molecular Differences: HPV Associated and HPV Unrelated

This translates into quite key molecular differences as well with HPV-associated disease characterized usually by Tp53 wild-type enriched for other mutations such as PIK3CA and others, APOBEC mutational signatures and really driven by the viral oncoprotein malignant transformation and progression process.

On the other hand, HPV-negative or HPV-unrelated disease is characterized by carcinogen-driven accumulation of somatic mutations and therefore is characterized by Tp53 mutations in the vast majority of these, as well as a number of other recurring mutational events, including EGFR and MET dysregulation, which we will talk more about as well.

LA HNSCC: Improving Outcomes

In the locoregionally advanced head and neck cancer setting, again, this is the setting where we see and treat most of our patients in the clinic. Our goals for these patients really vary by the biological subtype. Of course, HPV-positive oropharynx cancer, where survival is more favorable. Much of the efforts have been on therapeutic de-escalation to try to improve long-term function.

HPV-negative or HPV-unrelated disease continues to have quite modest survival, and therefore much efforts continue to pour in even in the immunotherapy era and improving outcome for these patients. Of course, as a medical oncologist, my interest is how can we incorporate or think about some of these promising new systemic therapies.

This image on the right is just to highlight some of the robust and fast responses that we can see with some of our active systemic therapeutic strategies.

Biologic Considerations for Immunotherapy in LA HNSCC

We have also learned quite a bit about the biological considerations for immunotherapy in locoregionally advanced head and neck cancer setting, and hopefully this will set the stage of what you will hear about subsequently from Dr. Haddad.

Of course, we have known for some time that immune checkpoint inhibitor therapy improves survival with or without chemotherapy in recurrent metastatic head and neck cancer setting, but it has been more of a challenge than many of us anticipated in terms of incorporating immunotherapy into the locoregionally advanced setting until quite recently.

This was initially reflective in a number of negative trials, where immunotherapy was tested in combination with standard fractionated, definitive chemoradiation. There is a number of different potential reasons why that may be. I have listed some of them here, including:

• The incorporation of both HPV-related and unrelated disease;
• The lack of biomarker selection in terms of PD-L1 expression;
• Concurrent immunotherapy administration; and
• Large field radiation to the neck.

Actually, Dr. Haddad published recently a very nice review, where he highlights some of these topics.

One of the common themes is that timing seems to be biologically important, favoring the neoadjuvant setting, actually not just in head and neck, but other malignancies as well.

Immunotherapy in Curative Intent: IO + Definitive Chemoradiation

This is just to highlight a number of the negative randomized trials of combining immunotherapy with standard fractionated, definitive radiation strategies for curative head and neck cancer. I will caveat here that while all these trials were negative, KEYNOTE-412 when enriched for PD-L1 expressors and with longer follow-up seemed to demonstrate a signal and there was a distant recurrence signal in the GORTEC-REACH trial. Nonetheless, these were negative trials and quite disappointing.

Optimizing IO With Multimodality Treatment in Curative Intent

There is a number of reasons for that. We have learned from a number of preclinical models a lot about the biology of head and neck cancer and immunotherapy that hopefully will set the stage for what Dr. Haddad will talk about, where we have understood that, in fact, the tumor draining lymph nodes of the neck are quite important for the activity of immune checkpoint inhibitors. We have seen a number of preclinical models. This is an example of one, again, highlighting that late ablation, in this case, surgical ablation of some of the neck nodes when testing with PD-1 or CTLA-4 immune checkpoint inhibitors seems to have more favorable responses.

Immunotherapy in LA HNSCC

This has led to the testing of immunotherapeutic strategies in the neoadjuvant setting. You will hear more about KEYNOTE-689 and some of the other trials. This is just to highlight that there is other strategies in development and of interest in terms of incorporating immunotherapy in combination in the neoadjuvant setting for surgical patients.

DEPEND: Induction Immunochemotherapy Study Design

This is just to highlight that there remains an unmet need in the definitive chemoradiation in nonsurgical patients with locoregionally advanced head and neck cancer, where we have yet an unmet need. You will hear about some of the ongoing phase III efforts in Dr. Haddad’s talk. This is one of the efforts that we have been very interested in, which is rather than giving immunotherapy concurrently with chemoradiation, rather giving it in the neoadjuvant and adjuvant setting and trying to give smaller volume radiation to the neck for some patients.

Induction Response to Immunochemotherapy in HPV-Unrelated HNSCC

Again, just to highlight that chemo-immunotherapy is quite an active regimen in previously untreated locoregionally advanced head and neck cancer with promising survival.

Efficacy Outcomes With Induction Chemoimmunotherapy Followed by Response-Stratified CRT

We look forward to more phase III data for this very important and large cohort of patients with locoregionally advanced, in particular, HPV-negative definitive chemoradiation-based patients to improve outcomes.

R/M HNSCC: Survival After ICI and Platinum-Based CT (2L+)

The other unmet need setting, of course, is the recurrent metastatic setting, where even in the immunotherapy era, we see still quite poor survival. That includes not only in the frontline setting with immunotherapy and chemoimmunotherapy, but also in the second-line and later setting after patients progress on immunotherapy and platinum-based chemotherapy.

This was real-world data from ASCO that we presented, looking at over 2,000 patients that had previously received PD-1 and platinum-based chemotherapy. Interesting, only 40% of these patients actually went on to receive subsequent therapy. Of those patients, the survival was quite poor, both for HPV-positive as well as for HPV-unrelated but worse median survival for HPV-unrelated. You can see median survival only 6.8 months in the real-world. So really an area of unmet need where we need better and more innovative treatments.

EGFR Dysregulation in HNSCC

Just highlighting some of the biological targets that you will hear more about. EGFR dysregulation within head and neck cancer has been known for a long time, particularly in HPV-unrelated or in carcinogen-induced disease. We know that the mechanism of this relates to EGFR ligand activation, dimerization and subsequent transphosphorylation and downstream signaling.

We have had for some time, cetuximab, an EGFR inhibitor, as a standard approach across treatment settings, although we know that the activity of this is less than we would like with only modest anti-tumor activity related to mechanisms of resistance.

EGFR, c-MET, and TME in HNSCC

There is a number of mechanisms of resistance we think that contribute to EGFR inhibition alone. You can see some of these highlighted here. MET targeting, as you can see here, as well as HER2, HER3, IGF1R, as well as factors related to the tumor microenvironment.

We know that a number of immune suppressive targets, TGF-beta, among others, may play a key role as well in EGFR inhibition.

Role of Stem Cell Receptors (eg, LGR5) in Treatment Resistance

Recently, we have also seen interest in the biological underpinnings of stem cell receptors in head and neck cancer and how those might drive treatment resistance to EGFR inhibition. One of the stem cell receptors is LGR5, which is an epithelial stem cell receptor which promotes cellular proliferation through Wnt and beta-catenin signaling. We know that this is overexpressed in head neck in a subset of cells that have some stem-like features. We think that this may also be driving some of the resistance and is currently being targeted in ongoing clinical trials, as you will hear more about as well.

ADCs: A Promising Therapeutic Avenue for R/M HNSCC

Finally, in terms of the biological vulnerabilities of head and neck cancer, I would like to point out the fact that antibody drug conjugates, which are drugs that have an antibody targeting component along with a cytotoxic component, also has biological rationale in head and neck cancer, where, in head and neck cancer we have overexpression of multiple cell surface antigens, that this is a biology that is sensitive to cytotoxic payloads, that it actually is a permissive tumor microenvironment. There has been a number of targets that have been and are ongoing being tested in the context of antibody drug conjugates in head and neck cancer as well.

Unmet Need in Head and Neck Cancers

Hopefully, this highlights some of the unmet need in head and neck cancer. We see very poor survival, particularly in the recurrent/metastatic setting, in particular, after the progression and resistance to PD-1 and platinum-based chemotherapy. I hope I have impressed upon you the distinct biology between HPV-related and unrelated, and the need to develop drug targets and drug development that target some of these specific biologic therapeutic vulnerabilities.

We also have a need for improved biomarker-directed approaches beyond just PD-L1 and HPV in head and neck cancer. Again, multimodality treatment continues to be associated with substantial toxicity, and there remains an unmet need for treatment optimization that not only improves survival but also prioritize organ preservation and function.

Case: 49-Yr-Old Woman With Recurrent HNSCC

We will just go back to our lady that we treated. This was a patient who we enrolled on a trial that tested amivantamab, which you will hear more about. Just to highlight, for this particular lady, she had a very nice clinical and radiographic response. Just highlighting the unmet need for these patients, the need for better therapeutics.

Hopefully, this gives some context for what you will hear next from Dr. Haddad in the locoregionally advanced setting, as well as from Dr. Burtness in the recurrent metastatic setting.

With that, I will hand it over to Dr. Haddad. Thank you for your attention.

Curative-Intent Locoregional HNSCC: Evolving Multimodality Treatment

Dr. Robert Haddad (Harvard Medical School): Thank you, Ari. Great presentation as usual. Good evening. Thank you all for sticking around on a Tuesday afternoon. I really appreciate you being here this afternoon.

Case Study: Tongue Cancer

This is the case that we presented initially. This is a patient from my clinic, a young woman with tongue cancer. Typical story. These patients often have a low level of suspicion when they go to their dentist or primary care because they do not fit the profile of a head and neck cancer patient. This patient took some time before she was diagnosed with a squamous cell carcinoma of the oral tongue T3N0 disease. This is when she was referred to us.

DFCI presentation

As we showed before, she had a four centimeter mass on the left side, a depth of one centimeter. Tumor did not cross midline. It was a lateral tumor. No involvement of tongue base or floor of mouth. As you know, those of you who listened to me yesterday in one of the lectures that I have given, we are having to expedite CPS testing in some of these patients, because when they come to us, we do not have their tissue.

The biopsy was done in another institution. We are doing biopsies in the clinic, since I am lucky enough to work with five surgeons. We do multidisciplinary clinics five days a week at the Farber, so we always have a surgeon in the clinic with us. Actually, we have also four oral medicine docs with us who can do biopsies.

I live in an environment that is a little bit, I would say, privileged or that when I need an instant biopsy, I can get that done. We did this and a CPS was 25. To complete the staging workup, obtained a PET scan.

PET/CT: Pretreatment

Those are the PET scan from this patient. You see this is a lateral tumor, a deep tumor, no evidence of adenopathy that we see on these scans. The Tumor Board recommendation was for perioperative pembrolizumab, which is what most of you have picked when you have answered that question. This is the image of the tumor in the clinic.

JAVELIN Head and Neck 100: Avelumab + CRT vs Placebo + CRT in LA HNSCC

The story of how we got to the perioperative immunotherapy obviously is quite interesting. Those of us on this podium here have been involved with immunotherapy now for quite a long time, obviously starting with the recurrent metastatic leading to curative setting.

The curative setting is not new because we started really testing immunotherapy in the curative setting with this JAVELIN phase III trial that we initially were involved in asking a question of adding a checkpoint inhibitor in this situation, in this trial, a PD-L1 inhibitor called avelumab in a patient population that is being treated with chemoradiotherapy.

This is not a surgical population. We picked essentially advanced head and neck cancer. They got received cisplatin radiation, randomized to avelumab chemoradiation or placebo chemoradiation with a maintenance phase of the PD-L1 inhibitor and a primary endpoint of progression-free survival.

JAVELIN Head and Neck 100: PFS

This, unfortunately, was the first phase III study to not show a benefit of a checkpoint inhibitor added to chemoradiotherapy. Obviously a huge undertaking to do these phase III trials. They take a very long time but no benefit to the patients, unfortunately, with the addition of this agent.

KEYNOTE-412: Pembrolizumab + CRT vs Placebo + CRT in LA HNSCC

You heard from Ari about the KEYNOTE-412, again, a similar trial in a way unselected for CPS. It was not really selecting the patients that we think would benefit the most from a PD-1 inhibitor, but a similar design to JAVELIN, but this one with the PD-1 inhibitor, not a PD-L1 inhibitor. Here it is pembrolizumab with chemoradiation or chemoradiation.

Very large trial, more than 800 patient. Primary endpoint of event-free survival and a number of key secondary endpoints.

KEYNOTE-412: OS in ITT Population

As you heard, the top line data of the trial was a negative trial, even though subsequently on further analysis of the groups with the high CPS, there seems to be a benefit with the addition of pembrolizumab to radiation and chemotherapy. Albeit, not an FDA-approved regimen currently, many have adopted a variation of this and some of the patients perceived to be a very high-risk with a high CPS.

Something to really knowing the literature is important as you try to select which patients should receive what therapy in this era of checkpoint inhibition in head and neck cancer, knowing what we know about the value of CPS in these patients, especially those who walk to your clinic with a very high CPS score.

IMvoke010: Atezolizumab After Multimodal Definitive Therapy for High-Risk Locally Advanced HNSCC

The other study that also we were involved in early on is this adjuvant phase III trial, which we modelled after the lung cancer data that was evolving or emerging with durvalumab, another PD-L1 inhibitor from AstraZeneca, in the lung cancer space, showing that an adjuvant approach in lung cancer did improve outcome.

We modelled this ambitious trial early on looking at atezolizumab, which is also a PD-L1 inhibitor. Again, this was a relatively small study, 400 patients randomized to either atezolizumab or placebo. Again, the reason I say we were ambitious because we picked essentially any patient who would have received curative therapy. It could be a surgical patient or a chemo RT patient, could be someone who got induction chemo followed by chemoradiation, upfront chemoradiation or surgery, followed by chemoradiation.

We really went with a large group of patients. We just wanted them to be in remission, disease free, and then they were randomized. So a pure adjuvant concept.

IMvoke010: EFS and OS

Again, no benefit seen here. What I would tell you here when you look at the overall survival, and this is true for all the previous trials that I showed you, because I have been doing this for 25 years now. Every time you do these phase III trials, that control arm seems to be constantly getting better.

When we initially wrote the design of this trial, these are not the overall survival curves we expected for this patient population with very advanced head and neck cancer. Again, it reflects the notion that head and neck cancer treatment is improving. Patients are living longer. Obviously, we still need to do better because, remember, the patients enrolled on these trials were patients with poor prognosis, very advanced head and neck cancer, all of them, the JAVELIN, the 412 and this trial.

Rationale for Perioperative Immunotherapy

No benefit with these. The first two were concurrent. The third one was an adjuvant, and leading me to talk a little bit about the perioperative space, which is fortunately where we have now a positive phase III trial.

As you heard from Ari, really, when you think of head and neck cancer, we think of HPV-positive and HPV-negative as two different distinct entities with different profiles, risk factors, prognosis, most importantly.

For the HPV-unrelated head and neck cancer patients, the outcomes continue to be relatively poor. We still cure maybe 50%, 60% of the patients. There is a need to do better. Larynx hypopharynx are historically difficult cancers to treat with a risk for distant relapse that is quite high, especially, for example, when you are dealing with oropharynx.

We know, as Barbara is going to discuss, that immunotherapy is a proven therapy for recurrent metastatic head and neck cancer patient. It improves outcome. I never imagined that I would practice in an era where I would have many patients in my clinic with complete response in metastatic head and neck cancer, who would be cured, who can stop treatment after two years and would be cured. This is not really something that I expected I would see in my lifetime.

The point is that these agents have improved overall survival in the recurrent metastatic setting. There is really a reason to think of why you would want to move these interventions to the perioperative neoadjuvant space, because these are active agents in head and neck cancer, just like they are active in lung and breast and kidney and we can go on and on.

Again, the unique aspect of the neoadjuvant approach is that you essentially are treating a patient when the patient is treatment naive, their immune system is intact. Unlike when you were doing concurrent chemoradiotherapy and the high impact of 70 Gray of radiation on the lymphocyte on the immune system, here you are essentially targeting a patient population that is treatment-naive. So this was appealing as a concept to target this approach of perioperative setting.

Obviously, in the back of our mind, every time we talk about neoadjuvant approaches, we are thinking maybe that would result in some less intense therapy down the road, meaning there will be some de-escalation that can occur if you are achieving a immune reduction or a RECIST criteria response, for example.

Those are bigger questions that will need much longer follow-up and different types of trials. The concept of reducing the intensity of the adjuvant therapy is always something that is in the back of our mind when we design these types of neoadjuvant interventions.

Mechanisms of Action: Perioperative Immunotherapy

Mechanistically, obviously we think of these approaches of the immune checkpoint blockade with the tumor antigen burden that is in place before you have done a neck dissection, before you have done 70 Gray of radiation, when the lymph nodes are still functioning and functional. Think of it as maybe, use the word “vaccination” that you are really inducing this immunity early on that will persist throughout the course of your subsequent therapy.

Neoadjuvant and Adjuvant Pembrolizumab in LA Head and Neck Cancer

KEYNOTE-689: Perioperative Pembrolizumab + SoC for Locally Advanced HNSCC

Really the first set of trials that we do not have time to show all of these. The way we got to KEYNOTE-689 is by doing a number of smaller studies early on to, first of all, show that it is safe to do. Every time you are dealing with a curative setting in head and neck cancer, the first question you have to answer before you get into efficacy and survival is, is it safe? Because remember, these patients are going to the operating room. They are having big operations. Many are going to need a flap. There will be radiation after.

You always be thinking, am I going to increase the perioperative toxicity, hospital stay, bleeding, flap failure? All of these questions were asked in smaller studies with pembrolizumab and with other agents. The first study we did was nivolumab plus ipilimumab before surgery, and we showed that that was safe.

Then Ravi Uppaluri, when he was in WashU working with Doug Atkins, they went in the pembrolizumab one cycle, pembrolizumab two cycles. There is a large number of trials that were done that showed that these interventions were safe. From a surgical perspective, you could deliver perioperative immunotherapy, still do the operation and not increase infection, surgical complications, bleeding.

The KEYNOTE-689 design you have probably seen ad nauseam by now. I do not have to really hash out. I am just going through some of the principles of how we got here. This was, in a way, a simple design. It is really perioperative immunotherapy or not. You either go directly to surgery or you start with two doses of pembrolizumab. Then you proceed to surgery. Then after the surgery, you look at the pathology report.

If you have high-risk features, extranodal extension, positive margin, you have to give chemo RT. You still give chemo RT. We added pembrolizumab or not. If you did not have the high-risk features, you will get radiation alone without cisplatin. The randomization there was radiation pembrolizumab or just radiation. In a way, it is a simple design that is really meant to test the question of what happens with this perioperative.

Now obviously if you are designing a trial today, maybe you would not design it the same way. You might ask a question, do you need the pembrolizumab at the back end? Do you need the pembrolizumab with radiation? All of these questions I am often asked and they are all reasonable questions.

At the time when we designed that trial, and when you design these trials, you are trying to go with the knowledge that you have at the time. That was the design that was felt to be the most appropriate at the time that we went with.

The primary endpoint of this trial is event-free survival. It is a primary endpoint that often gets criticized, even though it is an endpoint that is FDA approved. Many of the oncology trials have EFS as the primary endpoint. I keep saying to people, an event in head and neck cancer is a big deal. It translates into poor survival and poor quality of life and terrible pain and suffering. This is an appropriate endpoint for head and neck cancer curative trials.

That does not mean that we are not assessing the overall survival. That assessment is ongoing. The number of events have not been reached.

Another major endpoint of this trial was the major pathological response, because everybody in this room, including us, we were very interested in knowing what happens in the surgical specimen when you give two doses of pembrolizumab. We needed to know what is happening on that pathological response, and we defined mPR, again, very aggressively as more than 90% essentially tumor kill in the surgical specimen.

We had a whole group of pathologists who were helping us studying this, assessing this. So there is some consistency. This was all central review.

KEYNOTE-689: EFS in CPS ≥1 Population (Primary Endpoint)

The CPS more than one is where the FDA label is. Unfortunately or fortunately, that is for the future to debate. The FDA label comes at a CPS of one or more. Remember, the vast majority of the patients on this trial had a CPS of one or more. It was more than 96%.

Essentially, the FDA looked at the data, the EMA multiple jurisdictions, and essentially granted approval for this agent based on this data. These are the numbers as shown here.

In a curative setting head and neck cancer, you do not often see these types of differences between two treatment arms. This is for the CPS one or more, hazard ratio of 0.7, clinically significant and clinically relevant.

KEYNOTE-689: OS in CPS ≥10 Population (Key Secondary Endpoint)

For the CPS more than 10, also similar results here with an improvement with the perioperative treatment.

KEYNOTE-689: mPR (Key Secondary Endpoint)

The major pathological response, which is again, less than 10% of residual invasive carcinoma, both in the lymph nodes and in the primary, very ambitious endpoint. These are the numbers. Give or take around 10% of patients will have mPR.

This does not mean that if you have 20% of viable cancer, you did not actually do well, right? This is what is called mPR on that trial. What we showed later in some of the presentations that Bryan Bell did, that patients who had pathological response actually were a group of patients that performed better in that pembrolizumab arm.

The answer to the question of what was the mPR as defined on this trial is shown here around 10%.

KEYNOTE-689: EFS by Risk Subgroups (ITT Population)

This is the event-free survival. As you all know, there were less patients in the pembrolizumab arm that needed chemoradiotherapy. This notion of downstaging might be real. Based on this trial, we observed that there were less patients having a positive margin extranodal extension and thus needing chemoradiotherapy as compared to the patients who did not get pembrolizumab in perioperative. These are the numbers shown here.

NIVOPOSTOP: Adjuvant Nivolumab + SoC for Locally Advanced HNSCC

Now, that is essentially the perioperative. I know in this meeting there probably have been 15 presentations about KEYNOTE. We can answer questions later about it. That is what led to this trial and the top line data.

As I mentioned yesterday, there is an accepted manuscript in your journal, in the new head and neck journal of the surgical data of this trial. This is going to be a very important manuscript for the surgical community that is asked a lot of questions about the surgical outcomes. How many patients had disease progression during perioperative treatment but still proceeded to surgery with an R0 resection?

All of that will be presented. Bryan presented that information in the Palm Desert meeting. Now it is going to be soon in your mailboxes in a full manuscript.

Now, the other trial that really had also an impact on our field is the NIVOPOSTOP. This was a plenary session at ASCO. It is now published in a very high impact journal, Lancet. It is a different trial than 689.

We make a mistake by comparing these two trials. They are very different trials. In this trial, everyone got surgery first, no perioperative intervention. All that group of intermediate head and neck cancer is not part of this trial. This is a pure high-risk patient population. You have surgery, and on your pathology, you have high-risk features defined as positive margin, extranodal extension, multiple PNIs, multiple nodes. They randomized here to nivolumab/chemo RT or chemo RT.

I showed you in the beginning that adding a checkpoint inhibitor to chemo RT does not work, but that was not in a surgical setting. That was in a chemo RT setting. This is a surgical setting, meaning all these patients had surgery first. It is a different design for a trial.

680 patients also. By head and neck standard, this is a large trial.

NIVOPOSTOP: Disease-Free Survival (ITT)

Primary endpoint of disease-free survival. This is a positive trial. These are the numbers. The nivolumab/chemo RT did better than chemo RT. There were some tweaks to this. The primary endpoint by central review was actually a negative trial, so they had to go back and look at the investigator-assessed outcomes. That was a positive trend for nivolumab.

The major benefit of this intervention appears to be on locoregional control. With the 689 trial, it was on distant control. There are differences between the benefit of the perioperative versus the post-surgical intervention.

The overall survival on this trial also is not mature. It is not that the authors are hiding the OS, it is just not available yet. That is really important to keep in mind that the overall survival assessment of both of these trials is ongoing.

As far as we know, and I said this yesterday, I am not aware that this is a registration trial. I am not aware that BMS is planning to submit this trial for registration. To give this intervention, you will have to rely on whether the patient's insurance company will cover it. I do not think it has made it to guidelines yet to NCCN or ASCO. Sometimes it does help when there is a high impact manuscript to get insurance approval if you want to give nivolumab after chemoradiation.

As far as we know today, there is not a plan to submit this trial for registration in the US or in Europe.

NIVOPOSTOP: Cumulative Incidence of Locoregional Relapses

These are the numbers of locoregional relapses in both arms, again favoring the nivolumab arm as compared to the standard of care.

NIVOPOSTOP: OS (Descriptive)

Again, OS assessment is ongoing, but it is good to see that the nivolumab arm is on top. You do want to see that that the trends are favoring the nivolumab arm.

NIVOPOSTOP vs KEYNOTE-689

If you want to look at these two trials, I said they are different trials because there is [inaudible]. Even though they are all surgical patients, but these are very different concepts. We make a mistake by essentially saying that DFS here was this number and 689 it was this number. I do not think these should be compared in this fashion. These are very different trials.

CAMORAL: Perioperative Camrelizumab + Chemotherapy in Resectable LA HNSCC

Now where do we go from here? I always say when I talk about 689 that 689 is the first trial that essentially shows a positive outcome in that perioperative, but not likely to be the last trial to do that. It is a trial that we need to build on with new concepts. There are tons of concepts that are coming your way. The idea being that now that you have this mPR of 10%, you have this EFS, how can we do better?

In 689, it was mandatory that the surgical approach is the same, meaning the surgeons were required to do the same operation, whether the patient was going to have perioperative or not. They were not meant to be those adaptive surgery or changing the radiation based on the response. That opens up a whole set of investigation of how you can design trials in the future with those adaptation to the response, meaning doing different type of surgery or different type of radiation.

What people seem to now have fixated on is this notion of using chemo immunotherapy before surgery. This is an example from China called the CAMORAL trial. This is a checkpoint inhibitor PD-1 plus chemotherapy two cycles before surgery or just immediate surgery. So similar design to 689 resectable head and neck cancer patient population. Primary endpoint event-free survival.

This is not registration. This is a randomized phase II.

CAMORAL: EFS in ITT Population (Primary Endpoint)

Again, it is showing the signal that when you give chemotherapy with immunotherapy, you are going to have better outcomes. Thus, clearly here, you do not have to be a statistician. You can put not two fingers, five fingers between these two treatment arms to tell you that really the addition of the checkpoint inhibitor to chemo is better than no perioperative treatment. So really significant differences between the two treatment arms.

CAMORAL: OS in ITT Population

This is the overall survival. Again, a clear separation. Again, these are phase II. Decent size, 60 patients per arm. I would like to see the manuscript. I am not sure it was published yet. It is still in abstract form. Take this with what it means. It is an abstract. It is a good data. We would like to see a full manuscript that is peer reviewed and in a medical journal.

CAMORAL: LRFS and DMFS in ITT Population  

These are the other data of distant relapse and locoregional failure, again, favoring the immunotherapy arm.

CAMORAL: Pathologic Response

Most importantly, this is what we are going to show you here, which is we know this. We know this from TPF. I have been giving TPF for 25 years. When you give chemo to patients, you are going to shrink their tumor. It is not really something that is novel or innovative. We know this. We give TPF. Many of our patients actually get into CR when we give TPF.

When I joined Dana-Farber with Marshall Posner, we actually used to biopsy patients after two cycles of TPF before they started their radiation. We wrote a paper a long time ago, we had a CR rate of 93% in the primary, not in the neck. Those respond actually very differently.

When you add immunotherapy to chemotherapy, you are going to have a higher rate of pathological CR and mPRs, which means the surgeons will be going to the OR with a tumor that might have been four centimeter and now it is one centimeter and having to decide what type of surgery I am going to do now. It might need to be a different surgery. That is what we like to do in clinical trials to ask these questions and hopefully answer them.

AK112-II-020: Neoadjuvant Ivonescimab (AK112) + AP Regimen for Resectable LA HNSCC

If you were at ASCO this year, you would have seen this trial. This is a bispecific antibody. This is another combination of a bispecific antibody with chemotherapy. This is given before surgery and then you proceed with surgery. Then based on the surgery, you get chemo RT or RT alone.

AK112-II-020: pCR and 2-Yr EFS

Here, actually, you are going to see something even significantly better than the one I showed you before. If you look at the overall pathological CR, it is now 53%, mPR is 78%. These numbers now are going up and up and up. Remember with 689, I showed you 10% of mPR. That is phase III.

Here, this is a smaller study, again also from China. This is ASCO this year. This is recent mPR, 78%. These are massive number of responses.

In this trial, the investigators did actually pursue a dose-adaptive surgery. They did a different operation based on the response. If you needed a total laryngectomy for example for a T4 larynx and you had a response like this, you would do then a partial laryngectomy on that patient.

AK112-II-020: Response

The overall response, if you take everything, it is actually 97%. Again, very promising intervention. We would like to see bigger trials confirming this benefit. This benefit is real, but what we cannot answer with these small trials is obviously the downstream effect of these interventions on the immune system. Are you losing something by adding chemotherapy early on? That is why we have to do these trials and ask these questions and hopefully answer them.

AK112-II-020: Predictive Biomarkers

There seems to be a benefit in the higher CPS groups with a more deep response, both in the primary lesions and in the lymph nodes.

eVOLVE-HNSCC: Volrustomig as Sequential Therapy for Unresected LA HNSCC After Definitive Concurrent CRT

Now, what the field also is looking at right now, we are still looking at the adjuvant setting, but in a more defined setup, which is the chemo RT patients. Remember all the initial trials I showed you: JAVELIN, 412.

We are now looking at specifically the patients who are getting chemoradiotherapy getting to a CR and then randomized to an adjuvant intervention. There are two phase III trials asking that important questions.

The first one is this one. It is an AstraZeneca drug with a drug that is a bispecific CTL-4, PD-1, volrustomig versus observation. Primary endpoint also progression-free survival.

JADE: Sequential Dostarlimab Post CRT in Patients With Unresected LA HNSCC

The second trial is called JADE from GSK looking at their PD-1 inhibitor compared to placebo, also in patients who have unresected head and neck cancer treated with chemoradiotherapy, also a large trial and actually way advance in accrual.

Hopefully, in the next few years, we will see results of these. The last two standing in the adjuvant space, because all the efforts now seems to be moving into perioperative.

General Guidelines for Managing Immune-Related AEs

I was also asked to talk about immune-related side effects. For those of you who are interested, these are things we do all the time now in oncology. We grade the toxicity of the checkpoint inhibitors by:

• Grade 1, which is asymptomatic to mild;
• Grade 2, it is moderate;
• Grade 3 is more significant, requires intervention;
• Grade 4 is life-threatening, often requires admission and drug therapy.

Endocrine irAEs

When we talk about immune-related adverse events, they are not all created equal. Endocrine side effects like hypothyroid or hyperthyroid are very common. They are very easy to treat. We all manage them in our clinic. We do not often have to get endocrine involvement, versus when you get into things like myocarditis or hepatitis or colitis, you might be getting into more consultative services to do maybe a colonoscopy or an echocardiogram or other types of interventions when you have advanced type of toxicities involving those organs.

They are not all created equal, and the interventions obviously are quite different. They often are based on steroids, which seems to help a lot with these patients. Again, some of them are more concerning than others.

Management of Immune-Mediated Pneumonitis

Here I show you some of the recommendations, suggestions of how you can manage these side effects from grade 1 to grade 4. Again, making the point that for the grade 4, you have to stop the treatment. You cannot really treat someone with grade 4 toxicities. The grade 2s and 3s, you sometimes can rechallenge the patient depending on the particular situation and what type of toxicity you are dealing with.

Pre vs Post Treatment

This is my patient that I started with. On the far left is how she essentially presented. Then after one cycle of pembrolizumab and then two cycles of pembrolizumab, she then proceeded to have surgery. On the final pathology, she was still a T3N0 and continued with the recommended radiation treatment afterward.

Thank you very much. I am going to pass the podium to Dr. Burtness.

Recurrent/Metastatic HNSCC: Treatment Sequencing and Emerging Therapeutic Innovation         

Dr. Barbara Burtness (Yale School of Medicine): Okay. He has not left me much to say, has he? Welcome to those in the room and those online.

I am going to talk about recurrent/metastatic disease. Here we are at a surgical meeting. The point that Robert made earlier that the potential for the potency of pembrolizumab in the perioperative setting was recognized because of studies that had been done in recurrent/metastatic disease that showed activity.

Our hope is that as we introduce some of the newer agents that are beginning to show activity in the recurrent/metastatic setting that we may be beginning to envision the neoadjuvant approaches of the future.

Outcomes for R/M HNSCC With Pembrolizumab

Just to set the benchmarks, what are the outcomes for recurrent/metastatic disease when treated with pembrolizumab? I will say these are patients with recurrent or metastatic disease that is not amenable to salvage resection.

If you have a patient whose tumor expresses PD-L1 with a combined positive score of one or higher, based on the KEYNOTE-048 trial, the objective response rate to pembrolizumab monotherapy would be expected to be 19%, with a median overall survival of 12.3 months.

For a regimen that was really a novel regimen for us in head and neck cancer at the time 048 was launched, pembrolizumab plus doublet chemotherapy. The way it was done in 048 was a platinating agent in 5-FU, there an objective response rate of 36% and a median overall survival of 13.6 months.

Looking at some real-world evidence, and this is not surprising because it is also been recognized in the chemo era. There is some suggestion that the patients with HPV-negative cancer have a somewhat less favorable picture than those with HPV-associated disease. Here anti-PD-1 therapy median overall survival of 11.5 months. The combination of anti-PD-1 plus chemotherapy, 12.5 months. This is significantly improved compared to the reference regimen.

KEYNOTE-048: Pembrolizumab ± Chemotherapy vs EXTREME in R/M HNSCC

This just shows you the design of KEYNOTE-048. Patients were eligible who had unresectable, recurrent or metastatic or both disease of the oropharynx, oral cavity, hypopharynx or larynx. They had to have good performance status. Tissue had to be available for PD-L1 testing. If it was an oropharyngeal cancer, it needed to be of known p16 status.

882 patients were randomized 1:1:1. They were stratified by PD-L1 expression and there the data at the time we were heading into this trial were most significant for patients with a tumor proportion score of 50% or higher. It is actually not a very large proportion of the overall patients, but they were stratified on PD-L1 expression, p16 status, ECOG performance status.

The control arm was what the FDA had been requiring for a long time, doublet chemotherapy plus cetuximab. After six cycles, patients went to maintenance cetuximab. The experimental arms were either pembrolizumab alone or pembrolizumab with the same chemotherapy but not with cetuximab. Again, chemo capped at six cycles, followed by pembrolizumab maintenance.

KEYNOTE-048: OS for Pembrolizumab Alone vs Cetuximab/Chemotherapy

This shows you now the data for pembrolizumab alone compared with the control arm. For those of you in the room, on the left you see CPS one or greater. The hazard ratio here was 0.74. If you looked at median survival, it went from 10.4 to 12.3 months. If you looked at four year results, four-year survival went from 6% to almost 17%.

For CPS 20 patients, the hazard ratio was 0.61. Median overall survival went from 10.8 to 14.9 months. If you looked at four year results, from 8% to over 21% of patients were alive at four years.

If you look at all comers, here also, it looked as if pembrolizumab alone was non-inferior to the historical regimen. Actually, in a subsequent publication, we pulled out those patients with a CPS less than one, and they do not seem to be deriving any benefit from the use of pembrolizumab. And so they should not be offered pembrolizumab monotherapy.

KEYNOTE-048: OS for Pembrolizumab/Chemotherapy vs Cetuximab/Chemotherapy

Here we have pembrolizumab plus chemotherapy, again compared to the control arm. Here you see a hazard ratio for the CPS one of 0.64, taking the median survival from 10.6 to 13.6 months and four-year survival going from 4% to almost 22%.

For the CPS 20 patients, hazard ratio, 0.62. Pretty similar difference in OS from 11.1 to 14.7 months here, over a 22% difference, or just a 22% difference in four-year survival. For all comers, 0.71, and four-year survival going from 4.5% to 19.4%.

When we did that subset analysis for the CPS less than one patients, they did benefit from the use of pembrolizumab plus chemotherapy, although they did not benefit from pembrolizumab monotherapy.

Novel Therapeutic Approaches for R/M HNSCC

This has been the standard of care since about 2019. We would love to see new things come into the mix. There are a number of targeted therapies that have been reported or are currently being studied.

Zanzalintinib is a multi-targeted kinase inhibitor, similar to cabozantinib that is being studied in the STELLAR 305 study. Ficlatuzumab is a monoclonal antibody to HGF, which is hepatocyte growth factor. We called scatter factor. That is the ligand ferment. When given together with cetuximab, we showed in a randomized phase II trial that for HPV-negative patients that had a pretty high response rate. There is now a second-line trial ongoing for cetuximab-naive recurrent/metastatic patients who failed platinum and immunotherapy to be randomized to cetuximab or ficlatuzumab plus cetuximab.

We have a number of antibody drug conjugates, as you were hearing about from Ari earlier. Tisotumab Vedotin. Enfortumab vedotin is interesting in that that target Nectin-4, is expressed on HPV-associated cancer as well as HPV-negative cancer. We are not seeing a ton of drug development in the HPV-positive space at the moment, but this is one that might have some promise there.

Sacituzumab govitecan has some activity. The B7-H3 targeted, ifinatamab deruxtecan. A MET-targeted agent, which we saw some activity at ASCO this this year. Corbus 701 is a newer generation Nectin-4 targeted agent. This also looks like it has activity in HPV-associated disease, and PYX-201 looks at a splice variant of fibronectin.

[01:00:45]

Novel Therapeutic Approaches for R/M HNSCC

We also have for the HPV-associated disease, some vaccines that are coming along. BioNTech 113 targets the HPV16, E6 and E7 oncoproteins. That is given together with pembrolizumab compared to pembrolizumab monotherapy.

CUE-101 was a pretty interesting compound that had IL-2 backpacks on top of the HPV16 vaccine. This agent has actually been discontinued its development and then some work with personalized vaccines.

Novel Therapeutic Approaches for R/M HNSCC

The big story that people have been particularly waiting for phase III results is in the space of EGFR bispecific antibodies. You heard from Ari in the beginning that we have been using cetuximab since probably 2000. This is a monoclonal antibody that is chimeric. It seems to be in the Goldilocks position in terms of its affinity for the receptor. It had an impact on overall survival, which is how it ended up in the control arm of all these studies. It was pretty modest.

As monotherapy, it had a response rate of about 12%. Because of the fact that between about 2000 and about now, so few new targeted therapies were coming into the space, there was actually gobs and gobs of research about what are the resistance mechanisms to EGFR inhibitors.

One of the things that turned out to be relatively common is that the more potently you were able to suppress epithelial signaling through the epidermal growth factor receptor, the more the cell was likely to adapt and begin to be more reliant on mesenchymal signaling programs. That is a process called partial epithelial to mesenchymal transition. It is driven by a number of mediators, a number of which turned out to be targetable.

It is also probably the case that the more along that partial EMT continuum you are, the more immunosuppressed your cancer may be, the more cancer-associated fibroblasts you may have. So agents which successfully target the cells’ adaptive response to EGFR inhibition might also be more immunogenic.

There are three compounds that are all in phase III study. The first is petosemtamab. This co-targets EGFR and LGR5. LGR5, as you heard from Ari, is really just expressed on stem cells. Its expression in previously untreated head and neck cancer is actually pretty low, but it appears that it upregulates as you inhibit EGFR. Of course, it is those stem cells that may be the most treatment resistant part of the tumor and might be most responsible for late failure or for the cancer cells that are not eliminated by immunotherapy or chemotherapy.

The phase III study here is the LiGeR-HN1, and it compares pembrolizumab alone to pembrolizumab plus petosemtamab.

Next is ficerafusp alfa. This co-targets EGFR and TGF-beta, which not only is a manifestation of EMT but is a potently immunosuppressive and immune excluding cytokine.

Then amivantamab, which co-targets EGFR and MET, and MET is the receptor for that HGF that I mentioned earlier. As you can imagine, when you are targeting the receptor, you get all those cells where the ligand is overexpressed plus those where the receptor is overexpressed.

We also have MCLA-129, which is a bispecific for EGFR and MET, and izalontamab, which is a bispecific for EGFR and HER3. I guess it is important to say that one of the other ways that EGFR resistance can emerge is that there is adaptive upregulation of signaling through other members of the EGFR family, like HER2 and HER3.

Emerging Agents in Phase III Clinical Trials: EGFR-Targeting Bispecific Antibodies

Also in the second-line, in addition to the FIERCE study with ficlatuzumab that I mentioned, we also have studies here with petosemtamab up against dealer's choice of best standard therapy for that line of therapy, and then a first-line phase III study with amivantamab. That is interesting because instead of just being up against pembrolizumab, here amivantamab and pembrolizumab are combined with a chemo drug, carboplatin, and that is put up against the KEYNOTE-048 standard.

OrigAMI-4: Amivantamab in R/M HNSCC

Let me just talk you through a little bit of the data with these agents. This is from the OrigAMI-4 trial. This was the first attempt to introduce amivantamab into patients with head and neck cancer. It had a number of cohorts looking at amivantamab alone or amivantamab in combination, amivantamab in HPV-associated disease and amivantamab in the perioperative setting.

We presented at ASCO, the results from an expanded cohort in the post PD-1 inhibitor and platinum chemotherapy setting.

OrigAMI-4: BICR-Assessed ORR and Best Response (Cohort 1)

This was 102 patients. The confirmed objective response rate here was 42%. Very high for such a previously heavily pre-treated population, 15% complete response. As you can see from the waterfall plot, a number of the responses were quite deep. As you can see from the spider plot, a number of these were quite early and quite durable.

OrigAMI-4: PFS, OS, and DoR (Cohort 1)

The median progression-free survival again in this second to third-line population, 6.8 months, median OS, 12.5 months. Median duration of response had not actually been reached at the time of the data cut, but 56% of responders had a response duration of six months or longer, and 63% of the responses were ongoing at the time of the data cut.

OrigAMI-4: Safety (Cohort 1)

Median treatment duration was just under six months, and the adverse events were not surprising. They were predominantly things that had been previously described with EGFR or MET inhibitors. People who also follow the lung cancer world will remember that when amivantamab used to be given intravenously, it had a pretty high infusion reaction rate.

This formulation used in the head and neck cancer trial is a subcutaneous formulation, which appears to be much more tolerable and much less likely to lead to an infusion reaction. So administration-related reactions reported in 15%. All of these either grade 1 or 2. There were no grade 3.

Treatment-related discontinuations were pretty low at 8%. Really no new safety signals. I will just call out that there are some things from MET inhibition that are a little different than what we are used to with EGFR inhibitors, so particularly hypoalbuminemia in 50% of the patients and peripheral edema in 24% of the patients.

OrigAMI-5: Amivantamab + SoC vs SoC Alone in R/M HNSCC

This is the design of the origAMI-5 study, which is currently accruing. It is an open-label study. Patients with HPV-negative, treatment-naïve, recurrent/metastatic head and neck cancer, who have not had a prior EGFR inhibitor are stratified by their performance status and their CPS. Then they are randomized to either the chemo/pembrolizumab regimen from KEYNOTE-048 or to a regimen where the 5-FU is supplanted by amivantamab in a subcutaneous fashion with co-primary endpoints of objective response rate and overall survival, and secondary endpoints as you see them there.

First-in-Human Study of Ficerafusp Alfa (Expansion Phase)

Let me show you some of the data with ficerafusp. This is the bispecific that gets EGFR and TGF-beta. These data come from a trial where ficerafusp was given 1,500 milligrams day one, day eight and day 15. There are other schedules and doses that have subsequently been evaluated.

Ficerafusp Alfa + Pembrolizumab in HPV-Negative R/M HNSCC With CPS ≥1: ORR and DoR

What they saw here was that there were responses in both HPV-associated and HPV-negative disease. The confirmed objective response rate for this combination was 54%. They had a complete response rate of 21% with a median duration of response of over 20 months. As you see there, 57% of patients who were responders had ongoing response at 18 months.

Ficerafusp Alfa + Pembrolizumab in HPV-Negative R/M HNSCC With CPS ≥1: OS

The overall survival, as you see here, 21.3 months. Very interestingly, not really different for the CPS one to 19 and the CPS 20 or greater. This is of particular interest because the KEYNOTE-048 data and we did not really talk much about CheckMate 651 today, but that combination of ipilimumab and nivolumab also seemed to have its greatest benefit in the CPS 20 population.

Ficerafusp Alfa + Pembrolizumab in HPV-Negative R/M HNSCC with CPS ≥1: Safety

In terms of safety, this appears to be a pretty tolerable agent. No treatment-related deaths. 93% of patients have treatment-related adverse events, but only 50% are grade 3 or higher. You see the usual EGFR-related adverse events and then some stomatitis, epistaxis and anemia, but in general, not an unexpected safety profile.

Petosemtamab + Pembrolizumab for 1L PD-L1–Positive R/M HNSCC

Petosemtamab. This is the LGR5, EGFR bispecific, the one that might target stem cells. Here, confirmed objective response rate of 63%. Unlike the other two agents here, we have a pretty substantial signal in HPV-associated disease at 50%. Here we do seem to be respecting that the high CPS patients do better. Here are the response rate, 47% in the CPS one to 19 versus 73% in the CPS 20 or higher, and 79% of patients alive at a year.

Petosemtamab in Advanced HNSCC

This shows you the waterfall plot. Most frequent adverse events, pretty similar. This is an agent that appears to have a higher infusion-related reaction rate with 21% of patients having high-grade infusion-related reactions.

This is moving forward in phase III, as we said.

LiGeR-HN2: Petosemtamab vs Investigator’s Choice in Previously Treated R/M HNSCC

This is the second-line study, LiGeR-HN2. Here, petosemtamab is put up against investigator's choice. Co-primary endpoints of overall survival and objective response rate.

How to Interpret Emerging R/M HNSCC Therapies in Practice

Just to sum up this part. It looks as if these novel EGFR bispecific antibodies offer genuine promise to overcome adaptive resistance to EGFR inhibition, and they might be what we have been waiting for a long time in terms of turning EGFR targeting into something more meaningful for our patients.

The agents that I have reviewed differ somewhat by toxicity profile and their relevance to HPV-related cancer. It may be that we will end up with several of these agents winning in phase III and being approved. It may be that it is not going to be one size fits all.

There are no data to indicate to-date whether any of these agents would preserve activity after progression on a different EGFR bispecific. That is an important question for us to answer as soon as we can.

The last point here is that overall survival really remains the gold standard in terms of new therapies for patients with recurrent metastatic disease. Objective response rate is really an important consideration, because as Robert was saying earlier, many of these patients are quite symptomatic with pain, with threatened airway, with difficulty swallowing.

Clinical Decision Points Revisited: Applying New Insights in HNSCC Management

Posttest 1

At this point, we can revisit our post-test questions. We will start with, which patient profile most closely reflects the population in which emerging EGFR/MET-directed strategies have demonstrated clinical activity in recurrent/metastatic head and neck squamous cell carcinoma?

1. HPV-positive disease progressing after definitive chemoradiation;
2. HPV-positive disease with prolonged response to PD-1 inhibition;
3. HPV status unknown but treatment-naïve; or
4. HPV-unrelated disease that is progressed after platinum and PD-1-directed therapy.

Vote your conscience, except you two.

All right. 100%. Wow. That is amazing. Great. I actually think that never happens. That is great.

Posttest 1: Rationale

The rationale here is that amivantamab and ficlatuzumab are both approaches that have been primarily evaluated in previously treated HPV-unrelated or HPV-negative disease. I will actually say with respect to ficlatuzumab, that is the first randomized, phase II study that did include HPV-associated and it was not active there.

Patient Case 1

This is the case that we have revisited a couple of times. 29-year-old female, non-smoker, who developed left tongue irritation. Observation was recommended. By a year later, she had a painful left tongue mass that you saw the photos. It was four centimeters. It did not cross the midline. The next CT was difficult to assess.

Biopsy showed that it was a squamous cell carcinoma, and her PD-L1 CPS was 25. She was referred for treatment.

Posttest 2

Which initial treatment strategy is most appropriate for this patient?

1. Induction cisplatin plus 5-FU followed by surgery;
2. Perioperative pembrolizumab plus surgery and risk-adapted postoperative radiation or chemoradiation;
3. Surgery followed by adjuvant pembrolizumab monotherapy; or
4. Surgery followed by observation irrespective of pathologic findings.

Okay. Perioperative pembrolizumab plus surgery and risk-adapted postoperative radiation or chemoradiation. I agree with that. Could we go back for one second just to address the surgery followed by adjuvant pembrolizumab monotherapy because that had a few takers. Now you can go forward again please.

Posttest 2: Rationale

You heard from Robert that there are data for using PD-1 inhibition in the post-operative setting for a patient who for one reason or another did not get neoadjuvant therapy. The most common thing there is you did not appreciate the extent of nodal disease. T2 tumor you did not think there was nodal involvement. Then you get to the surgery, and it is high-risk or there is a lot of perineural.

In that setting, the data that we have are from NIVOPOSTOP, where the patient would also get chemoradiation.

Patient Case 2

Patient case two. This is a 60-year-old woman who was treated a year ago with definitive chemoradiation for a right-sided T4N2, p16-positive oropharynx cancer, comes back with throat pain and a new cough. PET CT demonstrates increased uptake in the right base of tongue. She has approximately 10 lung nodules. The largest is 3.5 centimeters in the longest dimension.

Her NaVDx is 420 and her CPS is 20. She gets carboplatin/paclitaxel and pembrolizumab for six cycles and has a deep partial response on PET. Her NavDx has come from 420 to 18.

Posttest 3

Following clinical and radiographic response, what do you think the best course for this patient is after six cycles of chemotherapy and pembrolizumab? Would you:

1. Add amivantamab;
2. Continue chemotherapy without pembrolizumab;
3. Continue pembrolizumab without chemotherapy; or
4. Resect the area in the base of tongue that was involved by the recurrence since this was previously radiated.

Okay. According to KEYNOTE-048, this patient would have continued on pembrolizumab without chemotherapy. That is what 61.5% of you picked. Adding amivantamab is certainly intriguing. She had HPV-associated disease, where we think amivantamab has less of a signal.

Continue chemotherapy without pembrolizumab. That would be something you would consider if the patient had had high-grade toxicity from the immunotherapy, but otherwise not.

Posttest 3: Rationale

Okay. We can go on from there.

Posttest 4

This is another post-test question. For your patients receiving emerging EGFR/MET-targeted therapies for recurrent/metastatic disease, which monitoring and supportive care strategy do you find most appropriate?

1. Conduct skin, oral, nutritional, laboratory and symptom assessments, with early targeted referrals, if indicated;
2. Monitor skin and oral health routinely, but add nutrition and rehabilitation services only after functional decline;
3. Prioritize pulmonary, hepatic and endocrine surveillance and manage dermatologic toxicities as they arise; or
4. Use routine oncology monitoring and initiate specialist referrals for persistent grade 2 or higher toxicity.

Please vote.

Okay. Outstanding. To conduct skin, oral, nutritional, laboratory and symptom assessments with early targeted referrals, if indicated.

Posttest 4: Rationale

The rationale here is that we know that these EGFR/MET bispecifics or combinations can cause dermatologic oral metabolic and fluid-related toxicities. Monitoring these closely can allow you to intervene with a more protein rich diet, exercise strategy, improved nutrition, improved dermatologic management.

People are probably aware of this COCOON regimen, where you can pretreat patients with antibiotics and topical steroids, which has been used in the lung cancer setting. With amivantamab, we do not really have experience either in terms of how effective that is or how tolerable that is in the head and neck cancer patients. It was not required in origAMI-4, and so this hyper vigilant early intervention approach remains a very appropriate for our patients.

Additional Polling Questions 

Poll 3

These two polling questions are not really necessarily content based. Firstly, do you plan to make any changes in your clinical practice based on what you learned from us today?

Poll 4

Please take a moment to enter one key change you would plan to make in your clinical practice based on tonight's event? We do not show the answers to this. This is just a quality check.

Discussion and Audience Q&A

Now we have a few minutes for questions. Great. Thank you.

Dr. Haddad: We have a few minutes for questions. Some of you have submitted some questions. If you have other questions, please either come. We do not have a mic, but you can just shout and we will hear you. It is a small enough room.

There is a question submitted about how long after the last dose of pembrolizumab can surgery be done?

In our practice, we wait two to three weeks. We do this early on. As we see the initial consultation, we plan when is dose one going to happen, dose two, and we have a surgical date. Obviously that can be modified for a number of reasons that we had discussed extensively in this meeting. We tried to book that surgical date early on.

Remember, many of these patients need two surgeons. You want to be prepared. We spoke a lot about no delays, just be really systematic with this intervention and intervene early on.

Dr. Burtness: Can I comment about that? I do not think there is a safety issue to shortening that interval.

Dr. Haddad: Yes. Right.

Dr. Burtness: We like that interval because it gives more time for the T cells to migrate into the tumor and then leave the tumor again and establish a memory population against the tumor for long term. We have had a study at our institution where we take people to the OR 48 hours after a dose of an immune checkpoint inhibitor. That is perfectly safe, as far as we can tell.

Dr. Haddad: Thank you, Barbara. There is another question. Maybe, Ari, you want to take this one. When multiple EGFR-directed or EGFR/MET-directed therapies become available, what biomarkers or clinical features should guide the selection and the sequencing for patients after platinum and PD-1 therapy? If you have, let us say, many of these trials we are talking about are positive. How do you go about selecting?

Dr. Rosenberg: Yes, it is a good question. It looks like the question is specifically asking about after platinum and PD-1 therapy. Some of the EGFR-targeted therapies that Barbara spoke about are being tested in the frontline setting with pembrolizumab. This is a later line after platinum and PD-1 therapy. There, we have the amivantamab data, which as of today is investigational, but depending on how things move forward, may become available and clearly has quite an impressive activity for HPV-negative or HPV-unrelated head and neck cancer.

There is a number of phase III trials that we are still waiting on the results, which may also be able to inform things as well. We have the ficlatuzumab-cetuximab trial, which Barbara spoke about very elegantly, as well as the petosemtamab, LiGeR-HN2 trial. We will be very interested in looking at those results. Those are looking in the platinum and IO refractory settings, so later line after progression after platinum and PD-1.

We will have to look and see. Obviously, we are interested in other biomarkers such as MET and some of the targets and things like that, but those are all investigation, I would say as of today.

Dr. Burtness: One would be intrigued to think that a patient who progressed, let us say, on pembrolizumab/petosemtamab might be able to get amivantamab. It is interesting that in a way, all three of these bispecifics go after a similar pathway with EMT, but different mediators. I really think we are just going to need experimental evidence to know whether or not they have cross resistance or not.

Dr. Rosenberg: Yes.

Dr. Haddad: There is a question that is a little bit tricky to answer, but an important question, which is, which patients with HPV-related disease are good candidates for neoadjuvant therapy? Because as you know, on the 689, there was a very small percentage of patients with HPV-positive disease. Who wants to take this one? I have my own opinion.

I can start by saying, typically right now outside of a trial, we do not use neoadjuvant therapy for HPV-positive disease, because remember for those patients, often the questions are de-escalation questions less intense treatment, omitting chemo, less radiation, maybe surgery alone. There are a lot of questions being asked that all revolve around de-escalation. Neoadjuvant therapy question. Depending on what neoadjuvant you are thinking, is it chemo or is it something else?

I do though see a fair number of patients who have very advanced HPV-positive disease that are very symptomatic. For those patients, it would not be unusual for me to treat them, for example, with TPF induction. That is a really small group and the patient population that usually has good prognosis. Often we see them as stage I and we have a number of de-escalation trials for those patients.

It is not right now on our radar for neoadjuvant interventions, even though I do point out that HPV-positive disease is not to be taken lightly. These patients do have actually a good risk of distant metastatic disease, and it is still an area of active investigation. Barbara, maybe you can talk about the ECOG trial that is completed with nivolumab or about to complete, but other maybe approaches to this question.

Dr. Burtness: That was not a neoadjuvant trial. It was patients with higher risk HPV-associated disease, so T4 or smokers. They completed their chemoradiation and then they were randomized to nivolumab or not for a year. That study is completed. We are waiting for the outcome.

I do think there are patients who have high node number, extensive extranodal extension, T4 cancers. They remain high-risk and they are not the bulk of who walks through the door. We certainly see them.

At my institution, we are interested in why PD-1 inhibitor monotherapy is not as much better in HPV-positive disease. Initially, we all thought, “Oh, well, the viral antigens there, that is going to be a great target for immunotherapy.” It turns out there is a lot of immune silencing in HPV-associated cancers. So we are interested in demethylation and vaccine and other approaches to enhance the benefit of PD-1 inhibitors. We have some investigator-initiated neoadjuvant trials.

I had a case recently where I did chemo/IO in an HPV-positive patient for a very specific reason. It was a second cancer in a previously radiated field. It would have been a horrific surgery, would have been morbid to give radiation. So far, so good in terms of response, but we will see how the patient does long term. I do not know about you.

Dr. Haddad: You have done some work on neoadjuvant, Ari?

Dr. Rosenberg: Yes. No, I agree with the comments and I would just really echo Barbara's comments about that not all these patients do exceptionally well and there is an unmet need. Barbara talked in her talk about some of the HPV immunotherapeutic strategies or vaccine strategies that are in development. Those should be continued to be pursued. There is an unmet need for new therapies in HPV-positive, not just in the recurrent/metastatic setting, but also in the high-risk locoregionally advanced setting as well.