Paradigm Shift to Neoadjuvant Immunotherapy For Patients with Colorectal Cancers

In the inaugural event of the Global Cancer Movement, initiated by OncoDaily, Pashtoon Kasi, MD, MS, delves into pioneering advancements in colorectal cancer treatment. The virtual event, held from December 6-8, 2024, brought together experts to explore the transformative role of immunotherapy, particularly in mismatch repair (MMR)-deficient cancers, where remarkable responses have been observed.

Dr. Pashtoon Kasi is the Medical Director of GI Medical Oncology at City of Hope, Orange County. Previously, Dr. Kasi served as an Associate Professor of Medicine in the Division of Hematology and Oncology at the University of Iowa, following positions as an Assistant Professor of Medicine and Oncology at both the University of Iowa and Mayo Clinic. He also served as the Director of Colon Cancer Research and Director of Precision Medicine Research for Liquid Biopsies at Weill Cornell Medicine.

His research primarily focuses on “liquid biopsies,” particularly circulating tumor DNA (ctDNA) and circulating tumor cells (CTCs). Dr. Kasi is also recognized as an author and editor of several significant books, including “Research: What, Why, and How: A Treatise from Researchers to Researchers” and “Impact of Circulating Tumor DNA (ctDNA) in Patients with Gastrointestinal Malignancies.”

His scholarly contributions extend to over 100 publications in esteemed peer-reviewed journals. He also received the Yvonne Award 2024 by OncoDaily in the “challenging the status quo” category.

I’ll use colorectal cancer as the example here, but a lot of what we’ll talk about, I’ll show you some data across other tumor types as well, where, of course, some of it is the new novel drugs, but there’s also something to be said about the new adjuvant platform and what are the proposed mechanisms behind efficacy that we’ve never seen before, so we’ll go from there.

My disclosures are available here and also on our institutional and other ASCO websites. So I wanted to start off with the poster child for immunotherapy, which is mismatch of period deficient cancers in general, and in particular colorectal cancers, which are usually one of the most common cancers amongst endometrial cancers where mismatch period deficiency is seen, and how immunotherapy, the whole story evolved.

Some of you might know, but it’s a very intriguing trail worth highlighting that this is one of the first trials where the name pembrolizumab and some of the other drugs weren’t even used at the time, and one of the first few studies of anti-PD-1, anti-PD-L1 drugs were being conducted, and across as many phase one trials do, there were a number of patients with a number of tumors, and consistently and fortunately across GI malignancies, there was the constant tale of it not working while you saw activity in melanoma, renal cell carcinoma, other more quote-unquote heart or immunogenic cell types, patients with GI malignancies and in particular colorectal cancer weren’t responding at all, except for one patient on this particular trial.

As you can see, the patient had a CR, and this was one of, only one out of 30 plus patients at the time over the years that had this observation, which prompted the initial hypothesis that there might be something different about this patient’s tumor, and they were able to find the previous diagnostic specimens and were able to confirm mismatched repair deficiency.

This led to the subsequent trials where this hypothesis was formally tested, and as you can see, there is a significant difference between activity of single agent PD-1 in metastatic mismatched repair deficient colorectal cancer as well as, not surprisingly, in mismatched repair deficient non-colorectal cancer, there is activity as well, but none of the mismatched repair deficient colorectal cancer had anything in the realm of response.

This is also one of the few situations where the responses tend to be ongoing and durable, where for the first time we started hearing the word cure in patients with stage 4 metastatic refractory cancer patients, who at the time were patients who had progressed for patients with colorectal cancer after having seen multiple lines of therapy, so best supportive care, hospice, or trials were options for these patients for them to have ongoing durable responses, improvement in their performance status, and that being durable leading to cures was something that we had never seen before.

Again, moving from a little refractory setting to one of the examples of moving this therapy up the line, the keynote 177 was one of the key trials where pembrolizumab was studied against chemotherapy. Again, there is a proportion of patients which unfortunately still do not respond to immunotherapy, and some of it, if we really focus on the subset that don’t respond, is there is a handful of patients that might be a testing issue.

These patients, from a simplistic standpoint, to the patient and caregiver audience, the mismatched period deficiency that is tested by immunohistochemistry or the microsatellite instability, which is tested by PCR or NGS-based assays, or the tumor mutation burden or TMB high, these are different lenses of looking at the same question of these tumors being more immunogenic because they are hypermutated.

But despite that, there is still a proportion of patients that don’t respond in the metastatic setting. And again, as opposed to chemotherapy, which even if it works initially, then it stops working, the response is you keep seeing a plateau consistently in patients with MSI high, mismatched period deficient tumors. Not surprisingly, dual checkpoint blockade, you see slightly deeper plots.

The Checkmate 142 evaluated this in a refractory setting. And again, as you can see here, still there is a proportion of patients that didn’t respond to dual checkpoint blockade with a low dose of ipilimumab as the CTLA-4 agent alongside the PD-1 inhibition with nivolumab. And then again, this is one of the first markers that received agnostic approval because this mismatched period deficiency, as mentioned earlier, amongst colorectal cancer and endometrial cancer, where that’s the most common tumor type mismatched period deficiency.

If there is one biomarker that we should not be missing or is a subset where immunotherapy is not only something that leads to a response on the first scans, but it’s also something that leads to durable ongoing control and cures. So even in tumor types like pancreas cancer, we’ve published on a case series of a handful of patients that, again, if they were mismatched or period deficient MSI high, they respond to immunotherapy after having treatment in terms of chemotherapy and radiation fail for them. So in this paper and other papers, too, you’ll see reported mismatched period deficiency across multiple tumor types.

And in many institutions, this has led to reflexive IHC-based testing so that we don’t miss any of these tumors in GI as well as other malignancies. And again, you keep seeing that plateau after a period of time. Patients who have not progressed after 12 to 15 months on average, they continue to have an ongoing response to therapy.

And some of these patients, when they’ve gone for surgery, even if there is something radiographically visible, these patients end up having a pathologic response. So the question of even surgery or metastatic amino patient was having a checkpoint response is now a question at times, and it’s not unreasonable to watch and wait these patients. Again, the same study with epanova that I mentioned earlier showed that same plateau as well.

So the goal of immunotherapy here is not just the response here. It’s truly this plateau of patients where they don’t have progression, they’re alive years out, which was not the case previously. This plot gets better if we go from, as shown in this graph, to treated setting, refractory setting, to first-line NEVO-IPI.

This was tested in MSI high patient population. The study and the publication is highlighted in the reference below. Again, you see that the plateau occurs early and is durable.

So as the therapy is being moved up the lines of therapy, the responses tend to be improving over time, and the PFS and the data of patients who are disease-progression-free is getting better as well. This was just presented and published as well, and this is the first randomized look at checkpoint blockade. In this study, they didn’t have a NEVO-only arm as well.

The NEVO-only arm, the results are not out yet. It would be very important to know, do you need the additional help of CTLA-4 inhibition in this MSI high subset? But clearly, as you can see, the difference between chemotherapy versus immunotherapy, there’s a striking difference after that initial partial overlap in the first few months of, again, those subset of patients who don’t have a response. We have ongoing durable responses where the median progression-free survival has not even been reached yet.

So in terms of, you know, the data I showed you so far, it was all about metastatic or first-line and second-line or refractory. When we go to a new adjuvant setting, see the striking differences in the same subset of patients, in the same tumor types, with the same agents, and see how different they are. So the story started with the NISH-1 study.

This was the initial look presented and published in Nature Medicine in 2020. Highlighted in red are the mismatch repair deficient tumors. As you can see, for the most part, there’s not a single patient with mismatch repair deficiency who did not have a pathologic response.

Again, these are patients who are resectable, so going for surgery, so indicated here are not the radiographic responses. These are pathologic responses on the specimen and terminologies and things that you will see that I’ll highlight later is anybody having more than 50% regression in the specimen versus a 100% response or a PAT-CR, which is self-explanatory. And then you’ll see near-CR or major pathologic response, this line at 90% or more kill of the tumor or less than 10% residual tumor, vice versa, whichever language one uses in these different studies.

But I think it’s important to highlight we need some consistency in the language of identifying these responses. Some of these responses assessments date back to 1994, where responses were developed for patients with esophageal cancer undergoing chemotherapy and radiation. So how we’re utilizing similar version of some of those responses and then some reiterations for melanoma to our patients getting checkpoint blockade and colorectal cancer.

So I think it’s important to highlight that, you know, there might be differences that we need to consider. And, you know, the story got better with the Chalabi plot highlighted by some where, you know, you see the waterfall plot with almost every single patient, over 100 patient in each two study showed that you’re having the responses. So as opposed to the metastatic setting, where there was always a proportion that didn’t respond, not only are you seeing responses, but you’re seeing, you know, 100% pathologic complete responses, which is something that we had never seen before.

And then the NICH3 with the NEBO and the LAG3 inhibitor, again, very similar results. You know, you rarely see lack of response or progression, no progression for that matter. But more importantly, the goal is the people getting cure, which in the advanced or resectable setting is the patients who are cancer-free at three years.

And this is the same in the Nature Medicine paper that was published just a few weeks ago with LAG3 as well. With Dr. Cirsic and the colleagues at MSK, similar waterfall plots of every single patient responding, 100% disease progression or recurrence-free, as well as no need for chemotherapy or surgery. And follow-up data from the rectal cancer with dostarlimab showed similar results of ongoing durable responses.

And there was an update with the median follow-up of over 18 months that the patients continue to have and maintain 100% clinical complete response. We may have some more updates to see if there’s been any changes to this 100% story. But overall, you know, in a nutshell, what you see in the metastatic setting, it’s strikingly different in the neoadjuvant setting.

So, you know, what is the, you know, rationale or data behind it? One of the key randomized trials is actually in melanoma. This is a group of patients who got neoadjuvant and adjuvant versus adjuvant only. So, just a couple of doses of PD-1 alone, given the neoadjuvant setting, you can see clearly, despite the same drug, same patient population, same situation, there’s a clear difference of disease-free survival.

And the idea is that, you know, if you just take the patient to surgery without any neoadjuvant therapy, majority of the anti-tumor T-cells that were inhibited by PD-1 at these sites of disease, they’re all receptive. So, surgery is not only removing the tumor, it’s also removing your immune cells that are trying to respond to the insult, which is the in this case. So, adjuvant therapy is just relying on the existence of anti-tumor T-cells that are inhibited by PD-1 at these micrometastatic sites.

But the bulk of the army, so to speak, that was trying to fight against the cancer is now removed. As opposed to this, again, this is from the same paper that is highlighted here, it’s in the supplement. And for neoadjuvant, what is shown is the difference is if you administer neoadjuvant immunotherapy, in this case anti-PD-1 therapy, it’s intended to activate the large number of anti-tumor T-cells at the tumor site before the definitive surgery.

So, what ends up happening is the locally activated anti-tumor T-cells then go on to induce a systemic immune response, and more importantly, recognize the distant macrometastatic site, which lead to recurrence. You know, it’s the, as we say, it’s the enemy that we don’t see or the micrometastatic cells for which we often, quote-unquote, do adjuvant key mode immunotherapy that can work better because there would be a broader repertoire of immune cells activated. This is shown in a different graphic with the NEVO-IPI study that was presented recently at ASCO.

Again, the proposed rationale is, as you can see, compared to neoadjuvant versus adjuvant, you are able to establish a broader, many more broader repertoire of diverse immune T-cells, which are unfortunately removed alongside the tumor specimen if somebody just goes for surgery. And I highlighted here in some pastational work, you can see in the adjuvant versus adjuvant setting, the clones that are present at baseline and the number of fold that they were expanded, you know, you don’t need statistical tests to see the grossly visible differences between the two. And again, at the end of the day, the outcomes speak for itself.

So, the data here was for PD-1 alone. This is data that was just presented with a primary endpoint of event-free survival for a couple of doses of NEVO-IPI given before surgery. So, the amount of treatment and duration was the same, except for a couple of doses pre-op.

And you can clearly see this is almost a 30% difference in event-free survival, which is clinically significant and strikingly different. You know, so, kind of shifting gear, what about, you know, the cold tumors? Well, we were talking about melanoma, which is a hypermutated tumor from UV radiation. There’s data for lung cancer, which is, again, chronic inflammatory from the smoking.

And then MSI-HI is a totally different beast altogether from mismatch repair deficiencies. But the proficient tumors, which is in the metastatic setting, are only like 4%, and then the advanced setting are no more than 10% to 15%. You know, there are new agents that are showing activity in terms of next-generation CTLA-4 inhibitors or multi-immune activators with one of the drugs that we have had a chance to work with called botansilamab.

Based on the data of this showing some activity up to 24% in the metastatic setting and more so in the non-liver met situation. So, the same tumor, it’s interesting that the settings are making a difference. You saw the differences between metastatic and neoadjuvant.

There’s also a difference between the same patient having active liver metastases versus having no active liver metastases. So, the trial proceeded forward in the metastatic setting with the observation that the liver metastases patients were not deriving as much benefit. There’s something about the liver being a site of lack of immune surveillance or immune suppressive environments, higher number of Tregs, that the same immunotherapy wasn’t working as well.

There was some of the signal, but not as much as we saw in the patients who had no active liver metastases or lung metastases and similar outcomes in peritoneal metastases. But we took that observation as well as the observation from the neoadjuvant paradigm shift that’s happening across multiple tumors to show that in MSS tumors in our so-called NEST trial, we were able to show activity of the drug and the combo, not just in MSI-high, but also in MSS tumors.

And as we increased the duration of therapy, what we also noticed was, as we were talking about the pattern responses, even if there was a tumor present, often it was truly quote-unquote downstage, where if this was a T3N2 tumor, the nodes became negative and the final tumor was a T2N1 or a T2N0.

So, you went from a stage 3 down to a stage 1 cancer. So, you’re truly downstaging the tumors. And a downstaged immunotherapy-exposed patient is somebody who’s very different than somebody who has a response, as shown here in this cartoon, to chemotherapy and radiation, where historically pathologic response sometimes don’t equate to disease-free survival because, again, you’re not necessarily targeting micrometastatic disease with local therapy when it comes to radiation or surgery.

And for that matter, chemotherapy is not the same as giving somebody immunotherapy from an immune standpoint. Again, the signal persisted at ESMO GI where there was an update. And what was also striking was, as we increased the duration of immunotherapy, and that was also noticed in some of the niche studies, that as opposed to surgery at four weeks, surgery at eight weeks or so on average, you see some of those deep responses converting themselves into near CRs or CRs.

And, you know, highlighted here in a nutshell, if you look at neoadjuvant chemotherapy as a reference, as a platform to design trials in the Foxtrot group of investigators where, you know, 1,000 patients have been treated, chemotherapy only leads to a major pathologic response, no more than 8 to 9 or 10 percent. And this has been replicated in other studies from China and the Dutch working group. In the niche results for MSS, it was about in the 20s.

And as we progress from the next one to next two, this is going high in the 40s and 50s. Again, we’re talking about deep responders. In my opinion, the moderate responders, more than 50 percent responders are also patients who are probably driving benefit.

And that’s where there is also a big unmet need to kind of study this in more detail and have common definitions and also do similarly so we’re able to compare and contrast outcomes across studies that are being done globally. So, you know, I’ll leave you with some thoughts and challenges for future directions. And, you know, there’s a lot of, you know, as we’ve explored this field a little bit more, a lot of memory from previous studies of chemotherapy and radiation of pathologic response not being a good surrogate for disease-free survival.

I would like to challenge that argument because if you really look at some of the recent immunotherapy studies, this is the Nadina study, which I showed earlier, where they were looking at the new adjuvant immunotherapy, adjuvant immunotherapy. And what you see clearly is patients who had a path CR or near path CR, so-called major responses highlighted in green. These are patients who are 94, 95 percent cancer-free.

It’s a flat line, as you can see, over two years of follow-up there, as opposed to somebody who has no response. That person, you know, is in the red bar, and as I highlighted earlier, even the partial responses that are more than 50 percent, it is conferring a benefit that is, of course, more than just the person who’s not receiving response. So, you know, while this might have not equated in the era of traditional chemotherapy, radiation, and some of the other studies like breast cancer, this is in immunotherapy-exposed situations.

This is not the case. Again, this is the same looking at three groups, major, partial, pathologic, and 95, 76, 57, you know, you don’t need statistical calculations to see the striking differences between that being a good early surrogate for the disease’s survival at two years or three years, as we think about efficient ways of doing trials. And again, it’s not just melanoma and colon cancer, and this is a nice paper published in Nature Medicine, for chemoimmunotherapy or chemotherapy.

Again, they showed differences in the primary tumor having a major path response or no response. These are differences that are clearly visible. And what they also showed was, you know, it’s not just major 90 percent.

Who said 90 percent is 91 percent better than 85 percent? You know, what they showed was that each 1 percent residual viable tumor, if present or vice versa, if absent, led to an improvement in the hazard ratio. So, every inch of improvement that you can get in an era of immunotherapy is something that downstream would lead to a higher chance of being cancer-free. So, the unmet need is in MSS colon cancer, as a GI oncologist focusing on colorectal cancer, is the proportion of patients not cured by standard of care surgery and chemotherapy.

Again, these are MSS patients. And, you know, across studies, whether you look at the foxtrot with chemotherapy, the optical neocole, at least a third of the patients, if not more, especially if you really go into the advanced high-risk patients, you know, some of the disease-free survival drops down to even 40 percent, 50 percent for the high-risk T4N2 tumors. There’s an unmet need to help improve outcomes in these type of patients, and the additional biologics, like anti-VEGF, anti-GFR, and other agents have failed.

So, we have an opportunity to intervene and truly be a part of this paradigm shift by acting in the neoadjuvant setting. And in the neoadjuvant setting, compared to chemotherapy, immunotherapy is showing something that we never saw before. And even, you know, for those who question about the use of pathologic response, even with chemotherapy, the group at foxtrot, as well as the optical group, has shown that with increasing levels, so it’s a little difficult to see, but what is important to note is people who had a response even to chemotherapy, they are this flat line.

They did not have any progression at multiple years of follow-up, and this is not a small study. This is, you know, over a thousand patients in a crew, and so, and replicated by other tumor types, as well. So, it’s not difficult to say there is no tumor left behind, you know.

So, while you can have some false positive from a pathologic standpoint, if you started saying 10 percent response or 5 percent response, but it is not at all, in my opinion, as a non-pathologist, to tell that there’s little or no tumor left behind, and even for that matter, 50 percent or more is not something that you will have problem in terms of agreement and reliability of assessment. And then, for colon cancer, you know, there’s also confusions regarding if CT is helpful.

Some of the data is confusing when they combine MSI and MSS patients, because in MSI high patients, there is a robust nodal immune response that’s happening, so patients have a falsely identified lymph node that is not positive for cancer.

But in MSS, that’s not the case, and with good robust definitions, you can identify these patients with sensitivities as high as 80 to 90 percent. So, it’s not a tool that is not helpful and has some value. And in terms of just a couple of days ago, FDA issued guidance of using CTDNA for industry and other trials.

We can see that, you know, it’s a strong prognostic tool of DFS at two years, so patients who are MRD negative versus MRD positive. An early look in terms of a study that looks at pathologic response has CTDNA and MRD together can be something that leads to accelerated approval, which then, of course, needs to be corroborated by DFS, but the data, both for pathologic response and MRD positivity, negativity, cannot be ignored. So, with that, I’ll stop by, you know, I think there’s a lot of paradigm shift in data that’s happening across tumor types.

While we talked and focused about colorectal cancer as opposed to child with a MSI high mismatch repair deficiency story, but we’re seeing that is going to MSS tumor, and also that’s something that is a paradigm shift that’s happening across solid tumors.

So, while there is a need for more drugs, but also I think there is a need to recognize this early so that we can use this new adjuvant platform for drug development and also for a lot of proportional patients that are diagnosed with advanced but not metastatic cancer. The notion of surgery or adjuvant chemotherapy, is that the right platform to continue to be working on, or should we be pivoting to new adjuvant therapy? I think the data is hard to ignore.

My contact and social media handles are listed.

Advances were coming in these subsets of subsets. So, you had the 2% KRS-C23, which now has an option.

We saw the 4% HER2 have an option. Beta, finally we’re seeing some activity. So, there are some advances that are coming for subsets of subsets, and then in the refractory setting, you know, while we have a bunch of TKIs and agents approved, you know, they weren’t necessarily agents that were causing response.

Again, it’s a continuum of care, and we are getting individual drugs, you know, still helping the survival of patients years and longer. But I think it’s reassuring to see now the immunotherapy, the question, you know, so the notion of immunotherapy does not work for colon cancer is an incorrect statement moving forward. I would say the commercially available immunotherapy does not work for colorectal cancer, but with novel agents like botansilamab and other FC-enhanced drugs also looking into this space and also identifying that even in metastatic setting, there is the liver met versus non-liver met, which is not a small proportion of patients.

I think it’s healthy and promising to see activity of immunotherapy now making its way for patients with colorectal cancers, because, you know, target therapies, these are all exciting things, but I think if there is one type of therapy where cures lie, or where I started hearing the word cure, it was all immunotherapy responders. So it’s nice that it’s becoming an option for patients with colorectal cancers as well. I see.

So it’s mostly about immunotherapies and targeted therapies.”