HER2, otherwise known as ErbB2, is 1 of 4 members of the ERBB family. ERbB1, of course, is EGFR.

They basically exist as heterodimers, and they have to join together to form homodimers in response to ligand binding and downstream signaling. Basically every member of ErbB combined with every other one, but HER2 is always the preferred binding partner, as you can see right here, and that leads to all of the downstream signaling.

[00:12:59]

Tumor Agnostic Prevalence of HER2 Alterations (Mutations and Amplifications) in GI Cancers

Of course we know that there is a wide prevalence of alterations in HER2 across multiple tumor types, including the GI tract. For tonight, we are largely focused on amplification or overexpression. But certainly it is important to be aware that there are activating HER2 mutations as well, although those tend to be certainly uncommon, for example, in the stomach and for the large part, certainly no standard and not a lot of experimental implications at this point for the mutations. We really focus on overexpression and amplification.

[00:13:37]

Guideline Recommendations for HER2 Testing in Advanced/Metastatic GI Cancers

These are the guidelines for testing. Essentially, they are largely the same. I would correct an error in colon. Patients should be tested for HER2 status in the advanced setting irrespective of co-occurring mutations.

Certainly, as we just discussed, testing really happens for patients with locally advanced unresectable disease or metastatic disease for whom systemic therapy is warranted.

Now, in terms of the methodology of testing, we also heard a little bit. It is probably best established in gastric, where there is reflexive testing with IHC and FISH, which I will go into. Next-generation sequencing, of course, is also now considered standard of care. But really the primary methodology in gastric should be IHC and reflexive FISH testing.

In colon, it is an any and all approach. I was talking with Kanwal about this. Certainly, NGS is sensitive for the detection of HER2 amplification. But IHC would also be a quick test with turnaround.

Pancreatic, rare. So not something we will discuss. Again, in biliary tract cancer, Rachna, it sounds like next-generation sequencing is the one-stop-shop that you normally go with. And there is good concordance with IHC. Yes.

[00:15:06]

HER2 Immunohistochemistry Scoring: ToGA Study

This is the staining criteria. This is specifically the staining criteria in gastric. This is based established originally in the ToGA study. One thing to note, there is a specific staining pattern in gastric cancer that is called basolateral. It is not uniformly staining. The luminal aspect typically does not stain positive. Therefore, the staining occurs only on the basolateral aspects of the cell.

Unlike breast and actually unlike even colon, the staining is much more patchy and heterogeneous in esophagogastric than those diseases. For example, you will note that an IHC 3+ is really staining in 10% or more of tumor cells, which is a relatively low number.

[00:15:52]

HER2 Testing: Colon and Other Sites

Indeed, if you look at the testing in colon, the criteria for IHC 3+ is much higher. It is 50% or more for staining. Actually, 1 thing that I learned tonight for the first time is actually there is actually no standardized testing in biliary tract cancer and various cutoffs have been used. That actually makes it difficult to directly compare between studies.

Finally, as mentioned, there are also other mutations in HER2. These have varying incidence throughout the GI tract. Again, at this point, there is certainly no standard clinical implication for HER2 mutations when they occur rarely. There are few studies that are targeting these mutations in a tumor agnostic way.

[00:16:36]

Intratumoral Heterogeneity and Incomplete Membrane Staining of HER2

This slide really just speaks to the heterogeneity of HER2 expression in gastric, much more so than in a disease like breast cancer or even colon cancer.

[00:16:48]

HER2 Mutations vs HER2 Overexpression/Amplification in GI Cancers?

Finally, as discussed, there is a distinction between HER2 mutations and HER2 overexpression and amplification. Certainly what we are all advocating for in each of our disease states is testing for either overexpression or amplification. That is really the eligibility criteria and the basis for assessing response with the treatments that we will be discussing. Again, the role of HER2 mutations remains to be defined in these diseases.

[00:17:17]

         Poll 5

This is experts in the hot seat where you all get to choose what we talk about. Which of the following questions would you most like the experts to discuss?

1. How are you using HER2 testing in clinical practice;
2. What are the inherent challenges in HER2 testing for GI cancers and how can modern practices adapt to ensure the most accurate results;
3. Is there clinical rationale and/or evidence supporting HER2 low as an actionable marker for anti-HER2 therapy in GI cancers; and
4. What barriers have you faced with implementing, testing and interpreting results, and how did you address these barriers?

Please go ahead and pick yours.

The most popular choice is what are the inherent challenges in HER2 testing for GI cancers and how can modern practice adapt to ensure the most accurate results?

Maybe let me start with Rachna. Because again, I think 1 of the challenges of biliary tract cancer sometimes is limited material for testing.

Dr Shroff: Yes. That is the answer right there. In terms of challenges as it pertains to HER2 testing in biliary cancers, when you think about the prevalence of HER2 amplification, it is primarily gallbladder cancer and extrahepatic cholangiocarcinomas, though it can be seen in intrahepatic cholangiocarcinoma as well. But it is about 10% to 15% of gallbladder and about 15% or so of extrahepatic.

Those are the 2 out of the biliary cancers that can often prove challenging in terms of quantity of tissue available for comprehensive testing. Either it is biliary brushings or FNAs and specimens as such that will make it harder for us to do NGS as a whole, but HER2 as an example.

Then, as you already mentioned, there is not clear understanding around the defining criteria. And I saw it on the slide, but different trials have used different cutoffs. So the understanding of what we really mean by HER2 3+ is still a little bit of a cloudy area for us. The biggest issue is really the fact that we have limited tissue available to really ensure the ability to do the testing properly.

Dr Ku: The obvious follow-on question to that. In a patient for whom, despite best efforts, you just have insufficient tissue, have you ever used ctDNA as a surrogate?

Dr Shroff: Yes. Because again, when you are thinking about comprehensive biomarker testing, ctDNA is actually the concordance rate for things like IDH, for instance, is relatively high. As the platforms are getting better, the sensitivities are improving for a number of the other targetable alterations that have FDA approved drugs. Yes, we do reflex to ctDNA with the recognition that we are likely missing things.

Dr Ku: Got it. It really sounds like in your practice, at the end of the day, one way or the other, you make every effort to make sure that there is some form of biomarker testing.

Dr Shroff: To the point of repeat biopsy, sometimes if we are not able to really understand what the biomarker profile looks like.

Dr Ku: Got it. Kanwal, tissue is less of an issue in colon and upper GI.

Dr Raghav: Yes. One of the challenges, and I will just pull on that thread a little bit. Unlike mutations, amplification, which is a surrogate for overexpression is more of a continuum. It is not a binary component where either a mutation is present or absent. Therefore, amplification sensitivity in ctDNA can matter depending on how much tumor is actually being shed.

I do feel that tissue is the primary mode and all attempts should be made. Furthermore, the type of drugs that are approved, and we will be discussing some of this is, whether it is an antibody drug conjugate or an inhibitor, it depends on whether you are amplified or overexpressed vs overexpressed alone.

For example, T-DXd is actually approved in IHC 3+ population. You can see that there is a dramatic drop off both in biliary and colon cancer even when it is 2+ FISH positive. We just do not know what correlation there exists between ctDNA levels and that IHC expression.

There are some studies ongoing. But that is a key challenge. So I would always encourage getting some tissue.

Dr Ku: Yes. The situation in esophagogastric is pretty similar to colon. It is generally not a problem getting tissue. There are very, very specific circumstances. Someone with a retroperitoneal lymph node that is near vessels and inaccessible to biopsy, maybe someone with bone-only disease. For the most part, it is relatively easy to obtain tissue and certainly HER2, MMR, PD-L1 and Claudin 18.2 are all standard of care testing for esophagogastric.

[00:22:30]

HER2 in Gastric/GEJ Cancers

Dr Ku: Great. Then I will now turn to HER2 and esophageal gastric cancers.

[00:22:36]

          Poll 6

There is another polling question for instant insight. Maybe this time, we can try and see whether we can get some audience participation. In your current practice, when do you typically assess HER2 status in patients with gastric or gastroesophageal junction cancer?

1. At initial diagnosis for all patients with advanced or metastatic disease;
2. At diagnosis only if systemic therapy is being considered;
3. At disease progression on first-line chemo;
4. I do not routinely assess HER2 status in gastric GE junction cancer.

Please do not pick option 4. Okay, 50-50. At initial diagnosis for all patients with advanced or metastatic disease; at diagnosis only if systemic therapy is being considered. I actually think both of these options sound very similar. Certainly if someone's performance status does not permit for treatment, I do not know that biomarker testing will have a lot of relevance. Does anyone in the audience perhaps want to share their perspective on when they test?

Maybe I can ask a more pointed question. We talked about this a little bit. Do you only do testing for patients essentially with metastatic disease, or does anyone get biomarker testing in localized curative intent setting?

Maybe I can ask you over there. When do you get HER2 testing for gastric cancer patients?

Speaker: I responded first.

Dr Ku: Okay. Got it. Her response was that at an initial diagnosis for all patients with advanced or metastatic disease.

Dr Raghav: I would agree with that for metastatic and locally advanced disease, where systemic therapy is being considered.

Dr Ku: Right. Exactly.

Dr Shroff: It is a merger of the first 2.

Dr Ku: It is a merger of the first 2. Yes. Fair enough. Let us leave that as it is and move on.

Dr Raghav: The key thing there is, though, it has to be done at the time of initial diagnosis, not later. Both those choices had that.

Dr Ku: Clearly to say you never do a HER2-testing in gastric cancer, clearly would be an inappropriate answer. I do think also that the reason for having this information certainly available at the time you make treatment decision. In the modern era, certainly someone should not initiate first-line palliative therapy without, again, HER2, MMR, PD-L1 and Claudin 18.2 status.

[00:25:48]

HER2-Targeted Therapies in GI Cancers

This is a brief history of HER2-targeted therapies in GI cancers. I will focus specifically on gastric and go over the data.

Of course, it was in 2010 that we had an approval of trastuzumab for HER2-positive esophagogastric cancer. Really after that it was a long barren period of 11 years when unfortunately all the other treatments were ineffective. It was not until 2021 that trastuzumab deruxtecan, T-DXd, was approved in the second-line setting or third-line setting. And then actually the timeline actually goes off the slide. But yesterday we heard about the results of zanidatamab, and that is something that I will also discuss.

[00:26:30]

CAP-ASCP-ASCO Algorithm for HCPs: HER2 Testing and Clinical Decision Making in Gastroesophageal Adenocarcinoma

This is something I alluded to earlier on, but this should be the reflexive testing in gastric cancer where patients initially undergo testing with immunohistochemistry. If you have an IHC 3+ tumor that is positive and no additional testing is needed. On the other hand, if you have an IHC 0 or 1+ tumor, that is considered negative and no additional testing is needed.

FISH should be performed only when the tumor is IHC 2+ or indeterminate. On the basis of these results, then you should be able to initiate treatment.

[00:27:07]

ToGA: Chemotherapy ± Trastuzumab as First-line Therapy in HER2-Positive mGC

The ToGA study is something that all of us are familiar with. It has an important place in history in the sense that it was the first study that evaluated a targeted therapy in gastric cancer, and actually the first study that showed an improvement in overall survival beyond the psychological 1-year barrier.

Here you see the Kaplan-Meier curve in the intention-to-treat population. Now this is important because the study at the time enrolled patients on the basis of FISH. So any patient with a FISH-positive tumor was eligible.

[00:27:41]

ToGA: HER2 Subgroup Analysis

In a preplanned subgroup analysis, it became pretty clear that the majority of the benefit were accrued to patients who had either IHC 3+ tumors or IHC 2+ or FISH positive. At the time, how these data were treated is that in the EMA, it was really approved only based on the subgroup that had the most benefit.

The US actually initially approved it in either IHC 3+ or FISH-positive tumors. But we now know, of course, that the contemporary definition is IHC 3+ or IHC 2+ and FISH-positive. That is the eligibility criterion for all contemporary era studies of HER2 therapy.

[00:28:23]

KEYNOTE-811: First-line Trastuzumab and Chemotherapy ± Pembrolizumab in HER2+ Gastric/GEJ

We then again move on to the KEYNOTE-811 regimen. This was a study that evaluated the addition of pembrolizumab to trastuzumab and chemotherapy.

What is notable about this study is that approximately 85% of patients had tumors that were IHC 3+. At the same time, about 85% of patients had tumors that had a PD-L1 CPS of 1 or more.

The primary endpoint of the study ultimately was survival, but a key secondary endpoint was response rate.

[00:28:56]

KEYNOTE-811: Pembrolizumab/Trastuzumab/CT Antitumor Response at Interim Analysis 3

What is notable about this study as well was that it was actually the first study to be approved in the US based on an improvement in response rate only. Ultimately, in the final analysis, there was a 13% improvement in objective response rate with the addition of pembrolizumab to trastuzumab and chemotherapy. You will note that there was a significant proportion of patients who had a complete response, 17% vs 11%, who got trastuzumab chemo alone. Pembrolizumab also increased the duration of response to 11.3 from 9.5 months.

[00:29:31]

KEYNOTE-811: PFS and OS (3rd Interim Analysis)

This is the third interim analysis for PFS and overall survival. As you can see from the Kaplan-Meier curves, there is a modest benefit to the addition of pembrolizumab which pushes things out.

Now I would note that these are the curves in either all patients or in patients with a CPS of 1 or more. We do not see the data for the patients with CPS less than 1. In terms of overall survival, the hazard ratio actually was more than one for those patients. Now, I do not think that adding pembrolizumab to trastuzumab and chemotherapy for patients who are PD-L1 negative is harmful but there really is no benefit.

The current FDA approval for this combination is pembrolizumab and trastuzumab with chemo for PD-L1, CPS 1 or more tumors. But it is trastuzumab and chemotherapy alone for tumors that have PD-L1 CPS zero.

[00:30:25]

DESTINY-Gastric01: Trastuzumab Deruxtecan in Previously Treated, HER2+ Gastric/GEJ Adenocarcinoma

Subsequently, DESTINY-Gastric01 was a study that evaluated trastuzumab deruxtecan in the third-line setting. This actually was a randomized phase II study performed in East Asian patients that randomized patients to T-DXd vs a physician's choice of chemotherapy, either irinotecan or paclitaxel, for patients who had HER2-positive disease. Again, this was a strongly positive study.

The primary endpoint of the study actually was response rate, that was 51% vs 14%. If the response rate was positive, there was then a plan to analyze overall survival. Here also we see clear improvement in median overall survival 12.5 months in a third-line setting.

Subsequently, DESTINY-Gastric02 was a study that was performed in the US and Western Europe, and it was a second-line study. An important distinction between Gastric02 and Gastric01 is that Gastric02 also required a pretreatment biopsy prior to study enrollment, while Gastric01 allowed for archival tissue as far back as at the time of first-line treatment.

[00:31:34]

DESTINY-Gastric01 Biomarker Analysis: T-DXd in HER2+ Advanced Gastric/GEJ Cancer After ≥2 Prior Regimens

Now, this is important because subsequent biomarker work from DESTINY-Gastric01 really suggests that tumors that are clearly HER2-positive derive more benefit. One of the indirect pieces of evidence is that the response rate was slightly higher in patients who had a HER2-positive tumor confirmed after or during first-line treatment with trastuzumab vs patients whose testing occurred before they received first-line trastuzumab, and the response rate was 57% vs 49%.

The reason for this is, in the orange bubble on the right, we know through multiple data sets that up to 30% of patients with esophagogastric cancer actually lose HER2 positivity at the time that they progressed on first-line trastuzumab-based therapy. That is a mechanism of resistance, and therefore it is very important as much as possible to biopsy a progressive lesion on trastuzumab-based therapy to confirm that HER2 is still expressed. Because otherwise it would not make very much sense to offer an anti-HER2 therapy.

There were other data points in DESTINY-Gastric01. So higher tissue mRNA correlated with higher response rate. Then the detection of erbB2 amplification by Guardant360 in the blood also was associated with a higher response rate of 61% vs 34%.

[00:32:56]

DESTINY-Gastric04: Study Design

The most recent study, of course, is DESTINY-Gastric04, which was presented at ASCO last year. Gastric01 was in the third-line setting. Gastric02 was a single-arm study in the second-line setting. Gastric04 brought it all together. This was a global randomized study that looked in the second-line setting at T-DXd vs the standard control of ramucirumab-paclitaxel.

Again, an important eligibility criterion for this study is that patients had to have a pretreatment biopsy to confirm HER2-positive disease after progression on trastuzumab.

[00:33:33]

DESTINY-Gastric04: OS (Primary Endpoint)

As we all know, this was also a strongly positive study. In fact, the survival here is even higher than what was seen in DESTINY-Gastric02. The median overall survival was 14.7 months vs 11.4 months with the control arm. That was a difference of 3.3 months that was highly statistically significant.

[00:33:54]

DESTINY-Gastric04: Select Drug-Related TEAEs

This is something that we discuss in 1 of the pretest questions. Of course, the toxicity that all of us who give T-DXd have to be familiar with is the interstitial lung disease. In multiple solid tumor studies, we see an incidence of about 10%. In this case, this study actually was a little bit higher, 13.9%. There is a small rare chance of grade 5 toxicity. So fatal toxicities.

Now having said that, while interstitial lung disease thankfully occurs only in about 10% of patients, what is much more common are chemotherapy toxicities. Because at the end of the day, ADCs are cytotoxic chemotherapy drugs. Certainly, an important part of the management is management of myelosuppression and also the management of nausea and vomiting, which is an underappreciated toxicity. So patients really have to have strong antiemetic regimens.

[00:34:49]

HERIZON-GEA-01: Study Design

This brings us to the HERIZON-GEA-01 study. These results were presented several days ago by Elena Elimova. This was a phase III study that evaluated zanidatamab. Zanidatamab is a bispecific antibody, but more specifically a biparatopic antibody, which means that each arm binds to a different domain of the same molecule, in this case, HER2.

There have been 2 previous phase II studies with zanidatamab plus chemo, and zanidatamab plus chemo plus tislelizumab, which is an anti-PD-1 antibody which showed very promising results and that paved the way for this phase III study.

Again, the criterion for eligibility was IHC 3+ or IHC 2+ FISH-positive. That relates to 1 of the pretest questions. Patients with these criteria were randomized in a 1:1:1 fashion to these 3 arms. Arm A was trastuzumab and chemotherapy alone. Arm B was zanidatamab with chemotherapy, and arm C was zanidatamab plus tislelizumab plus chemotherapy.

All of us immediately appreciate that arm A, the control arm, excludes pembrolizumab which is the current standard of care. At the time that the study was designed and that was opened, the only place in the world where pembrolizumab was approved was the US. Therefore, this actually was an ex-US study, did not open at any US sites, but it did open in North America as well as the rest of Europe.

Dual primary endpoints of progression-free survival as well as overall survival. Key secondary endpoint was response rate.

[00:36:39]

HERIZON-GEA-01: PFS per BICR (Co-Primary Endpoint) Zanidatamab + Tislelizumab + CT

What we did not show is that there is actually a hierarchical testing that we initially started with PFS for arm C vs arm A. So arm C being the triplet combination. If that was positive, then we would then look at PFS for arm B vs arm A. If that was positive, we would then look at overall survival for arm C vs arm A. Then finally we would look at overall survival for arm B vs arm A.

There is a hierarchical testing. Each prior result had to be statistically significant to initiate the next testing. This actually was the first arm and outcome to be tested. It was progression-free survival per BICR, so Blinded Independent Central Review for arm C vs arm A. You can see that this was strikingly positive. The median PFS is 12.4 months vs 8.1 months, hazard ratio of 0.63. That does translate to an 18-month progression-free survival rate of 43.9%.

[00:37:55]

HERIZON-GEA-01: PFS per BICR (Co-Primary Endpoint) Zanidatamab + CT

Based on this, the next combination to be tested was arm B vs arm A. This also was statistically significant. In fact, the median numbers were very similar, 12.4 vs 8.1. But if you look at the 18-month progression-free survival rate, it is a little bit lower at 38%, suggesting that there may be some benefit of the tail end of the curve from the tislelizumab-containing arm.

[00:38:24]

HERIZON-GEA-01: OS (Co-Primary Endpoint) Zanidatamab + Tislelizumab + CT

We then looked at the overall survival comparison between arm C and arm A. Again, this is statistically significant. In fact, it led to a striking median overall survival of 26.4 months vs 19.2. That is a hazard ratio of 0.72. That means that it is a 2-year overall survival rate of 54.3% and a 30-month overall survival rate of 43.8%.

[00:38:51]

HERIZON-GEA-01: OS (Co-Primary Endpoint) Zanidatamab + CT

The fourth analysis was overall survival for arm B vs arm A. While this is numerically better, 24.4 vs 19.2, with a hazard ratio of 0.80, the p value just misses statistical significance at 0.0564.

Now, I would quickly point out that this is actually the first of several planned interim analyses, and the next interim analysis will occur sometime in the middle of this year. Given the strong trend, the hope is that, at that point, enough additional events will have accrued that there will be a statistically significant improvement in overall survival, even for arm B vs arm A.

[00:39:32]

HERIZON-GEA-01: Safety

Now in terms of safety. Again, busy slide looking at the treatment-related adverse events. There were more grade 3 adverse events in arm C, the triplet combination, that was 71.8% vs 59% in both other arms. Ultimately what that meant was that there was an increased rate of zanidatamab discontinuation in the 2 experimental arms, so it was 13.3% in arm C, 10.5% in arm B vs only 5.6% discontinuation rate of trastuzumab.

There were also some other adverse events of special interest. One thing to note is that zanidatamab is associated with infusion-related reactions in about 25% of patients. The vast majority of these are grade 1 and 2 and can be managed with premedication.

Actually, we do not have a slide for diarrhea. It is worth mentioning that there was a higher incidence of grade 3 diarrhea in the zanidatamab-containing arm. It was about 24% in arm C, about 20% in arm B compared to about 12% in the control arm. This is relevant because diarrhea was identified as an emergent toxicity even in the phase II studies, and as such, all patients receiving zanidatamab on the study received mandatory loperamide prophylaxis 4 mg BID for the first week.

Despite that, the diarrhea in zanidatamab arms typically developed around Day 7, and irrespective of the severity, typically resolved after 21 days. Certainly, as we move forward, and I think as many of us look forward to being able to offer this drug to our patients, the management of diarrhea is going to be something critical and will require patient education, as well as educating all of our medical teams.

[00:41:51]

         Posttest 1

Post-test question 1. This is a question that was previously addressed. Based on the phase III HERIZON-GEA-01 trial, which HER2 testing results would make a patient with newly diagnosed locally advanced or metastatic GEA eligible to receive zanidatamab in combination with chemotherapy and tislelizumab?

1. IHC 3+ only;
2. IHC 1+ or more;
3. IHC 3+ or IHC 2+ with ISH positivity;
4. IHC 3+ or IHC 2+ regardless of ISH status.

Okay, this is fantastic. I am more than thrilled and gratified. The correct answer absolutely is IHC 3+ or IHC 2+ with ISH positivity. That is great.

[00:42:56]

Poll 7

Okay, another polling question. Experts in the hot seat. Again, please vote for what you would like us to discuss:

1. Is there a preferred sequencing approach for HER2-targeting agents in the second-line setting;
2. What is the potential impact on the first-line GEA treatment landscape considering the positive readout of the HERIZON-GEA-01 study;
3. What thresholds best predict benefit from continued HER2 targeting;
4. How many HER2-targeted therapies address unmet needs in earlier lines of care;
5. Should HER2-targeted therapy take precedence over other targeted therapies.

One-third each. Maybe the in-person audience will pick. The 3 options are:

1. Is there a preferred sequencing approach for HER2-targeted agents in the second-line setting;
2. What is the potential impact on the first-line treatment landscape with HERIZON-GEA-01; and
3. Should HER2-targeted therapy take precedence over other targeted therapies.

I feel like there is some overlap between the last 2 choices. Maybe given just the recent presentation of HERIZON-GEA, it makes sense for us to discuss that. Maybe let me open it to my non-gastric cancer treating GI medical oncology colleagues.

Kanwal, why do not you start with you? Again, as a gastric cancer adjacent person, what is your perception of the HERIZON data?

Dr Raghav: Yes, it is very exciting data, especially after such a long hiatus in first-line setting. Overall survival of nearly 26 months. That buys it for me. Everyone with gastroesophageal cancer gets HER2 testing at diagnosis and gets this once it is approved.

Dr Ku: Yes. Rachna, what do you think?

Dr Shroff: Yes, I agree. I had the privilege of being the expert commentary on this study, actually, and that is exactly what I said. I said that this is changing the landscape for HER2-positive advanced upper GI malignancies with this data.

The median OS of 26.4 months is very exciting. What is really great is that when you look across all the efficacy endpoints looked at, there is response that is correlating with Blinded Independent Central Review of PFS and then overall survival, which obviously is the game changer.

Dr Ku: One thing we did not discuss actually in the subgroup analyses, something certainly that was surprising to me is that in the tislelizumab-containing arm, both PD-L1 positive and negative tumors seem to benefit.

Maybe this is a tricky question. First of all, overall survival in arm B, the zanidatamab plus chemo arm is still not mature. Clearly, both experimental arms were not meant to be directly compared to each other, although that is immediately the first thing that everyone will do.

We all agree that zanidatamab plus chemotherapy is a new standard. What are your thoughts about the role of tislelizumab? Is it critical? Should it be offered to patients? Are there patients that you would not offer it to?

Dr Raghav: Like you said, the study is a study and its eligibility criteria is its eligibility criteria, and PD-L1 was not one of those. We need more studies regarding whether IO is adding. As of now, I would give the benefit of doubt to the patients and they would get the treatment that gives you the maximum survival advantage.

Dr Ku: Yes. Again, 1 thing we also did not show is arm C. So the arm that got zanidatamab, tislelizumab and chemo, the median duration of response was 20.7 months, which is mind boggling that someone can be on treatment for almost 2 years as a median, meaning that there are some patients who are on it for even longer than that potentially.

Again, we normally think of immunotherapy giving you a benefit in terms of the tail at the end of the curve. Maybe that is what we are seeing here. Yes. Rachna, what do you think?

Dr Shroff: Yes, I agree. At the end of the day, we can sit and hypothesize about the scientific rationale and what is happening. The way that the study was designed, it was designed at a time before we really could integrate the KEYNOTE-811 data into the control component of this. I guess I would say, unless there are reasons that the patient could not receive tislelizumab or immunotherapy, I would say that the 3 component approach is the right way to go.

Dr Ku: Yes. No, I absolutely agree. All right. Perfect. Thank you so much.

[00:48:12]

HER2 in CRC

Dr Raghav: Thank you. Good evening, everyone. Moving on to a little bit lower in the GI tract. This is HER2 in CRC.

[00:48:27]

          Poll 8

We are going to start with that instant insights. This is a poll question. How confident are you in your ability to appropriately order and interpret molecular testing in colorectal cancer and use the results to guide treatment decisions?

Great. This is spread all over: moderately, slightly not confident. Well, hopefully we can get you to a very confident level at the end of this presentation. We are already there with gastroesophageal I am sure. My job here is to convince you that the same stands true for colorectal cancer.

[00:49:22]

HER2+ mCRC: A Distinct Subset

HER2 in colorectal cancer is, in some cases, similar to what you see in breast and gastric. In some cases the nuances are unique to colorectal cancer.

HER2 is a low prevalence biomarker, but it is a distinct subset in CRC. It is about 2% to 4% of all unselected CRC. When you drill it down, most of this is enriched in RAS/BRAF wild-type patients. They are more likely to be left-sided and distal colorectal cancer. They are more likely to have lung metastases, and they also have a higher incidence of brain metastases.

A classic colorectal cancer, your brain metastases is about 2%. But in HER2, the prevalence is as high as 20%. So a lot of these patients do actually require brain imaging prior to beginning therapy as part of their restaging, which is not the conventional standard for colorectal cancer.

HER2 amplification is enriched in RAS/BRAF wild-type tumors, and this is a very critical point because in RAS/BRAF wild-type tumors, the current standard of care for colorectal is use of anti-EGFR antibodies, and HER2 amplification actually acts as a negative predictive biomarker for that efficacy.

The prognostic role is still unclear. This is not like BRAF, which we clearly know is a negative prognostic marker in CRC. We still do not know whether HER2 is a prognostic biomarker, but there is definitely a negative predictive effect in response to anti-EGFR.

[00:50:58]

CRC: A Molecularly Diverse Entity

This is the larger picture of colorectal cancer. You can see HER2 sits right there in the middle. Majority of it, 80% of all HER2-amplified tumors will be RAS/BRAF wild-type. Only about 20% are RAS-mutant. There are some quantitative differences in the kind of amplification and the levels of amplification that are found across those 2 boards. I want you to keep that in mind because this affects treatment decisions in colorectal cancer, which do not affect any other tumors outside of this disease.

[00:51:32]

HER2: Overexpressed/Amplified in 2% to 4% of CRC

HER2-expressed colorectal cancer is seen in 2% to 4% of all CRC by IHC/ISH. Geoff mentioned this that we can use IHC/ISH, as well as NGS. There are some nuances to that which we will discuss in the following slides.

[00:51:47]

HER2-Targeted Therapies in GI Cancers

This is the history of HER2-targeted therapies in GI cancers. You can see we started way later. The trastuzumab-tucatinib combination got approved in 2023. Trastuzumab deruxtecan, which is the antibody drug conjugate got approved in 2024 for IHC 3+ tumors.

[00:52:09]

MyPathway: Pertuzumab and Trastuzumab for HER2-Amplified mCRC

This was one of our first foundation studies for HER2-amplified colorectal cancer. This laid the foundation of how we treat colorectal cancer today. Just remember, this was trastuzumab and pertuzumab, so this is dual anti-HER2 therapy. You need 2 agents in HER2-amplified colorectal cancer unlike gastric cancer.

What is key here is this was a highly refractory setting. Despite that, trastuzumab and pertuzumab had good response rates. What is important to note is the KRAS status. When you have a KRAS-mutated patient, this does not work. The reason why this does not work is because KRAS mutation sits downstream of the HER2 signaling, so blocking the HER2 signaling does not really affect that.

Dual anti-HER2 therapy after this was used in HER2-amplified colorectal cancer only for RAS and BRAF wild-type patients.

[00:53:04]

MOUNTAINEER: Tucatinib + Trastuzumab for CT-Refractory, HER2-Positive, RAS Wild-Type mCRC

This was MOUNTAINEER study, which was again dual anti-HER2 therapy, just a different type of HER2 therapy with trastuzumab and a TKI called tucatinib. This is the first study that established an FDA approval for HER2-targeted therapy in colorectal cancer. As you can see, the eligibility criteria for this was HER2-positive RAS wild-type metastatic colorectal cancer. Because we had already learned through MyPathway that dual anti-HER2 does not work in RAS-mutant patients.

This was again a treatment refractory population, 117 patients. It started with a single-arm study which was trastuzumab and tucatinib. Then there was a randomization to trastuzumab tucatinib vs tucatinib for proof of component. This was an important learning lesson for us.

The primary endpoint was confirmed overall response rate by Blinded Independent Central Review.

[00:54:01]

MOUNTAINEER: Efficacy With Tucatinib + Trastuzumab

These are some of the results. I wanted to concentrate on the trastuzumab-tucatinib arm which is A plus B, 84 patients treated and a response rate of nearly 37%. This is a treatment refractory colorectal setting where the current standard of care response rates range anywhere between 2% to 5%. This is an immense improvement over what we would have otherwise given these patients. You can see 38% response rate with a very long duration of response.

One other thing to note from this study is when you give tucatinib alone, your response rate is 3%, which means that single-agent tucatinib or single-agent trastuzumab or single agent anti-HER2 therapy is not really effective in this disease.

Furthermore, patients who received tucatinib, when they progress, they were again given trastuzumab and tucatinib as a crossover, and the response rate never reaches what it was when you initially challenged them. So it is very important to identify these patients and actually treat them with dual anti-HER2 therapy.

The waterfall plot is also very revealing. When you look at the primary site of disease, you can see the left-sided disease tends to actually benefit a lot more. The response rates are higher, about 42% compared to 15% when you look at non left-sided disease. This is 1 of the unique characteristics of colorectal cancer. This is also seen in anti-EGFR.

Whenever you block this HER family of receptors, for some reason the left-sided tumors respond much better. When you look at other markers, the response rates are pretty uniform.

[00:55:48]

MOUNTAINEER: Final Efficacy Analysis

This is the final efficacy analysis that was presented. Again, the response rate stayed at about 39%. Look at that duration of response is 15 months of response which is not seen with even classic conventional chemotherapy in colorectal cancer. The median PFS was about 8.1 months and a median overall survival of 23.9 months.

One of the lessons that we did learn from the translational aspect of this study is whether you use tissue IHC or ISH or you use tissue NGS, you can see the response rates are pretty comparable 41% for tissue IHC and ISH. For tissue NGS, it is about 50%. Even if you use blood NGS, it was about 42.4%. Some of these differences arise from the fact that your NGS often picks up higher level amplifications.

[00:56:38]

MOUNTAINEER: Safety

This is the safety analysis. Not only was this therapy effective, but it was also very well tolerated. Grade 3 adverse events occurred in 40% of patients, but those that led to discontinuation was only about 5.8%.

Then the most common adverse event you can see is diarrhea. Majority of this is grade 1 and grade 2, but there are a few cases of grade 3 diarrhea and then some hypertension, fatigue. So very much class side effects that you would expect out of TKI.

[00:57:14]

DESTINY-CRC01: Trastuzumab Deruxtecan in HER2-Amplified mCRC

Now coming to DESTINY-CRC01, trastuzumab deruxtecan has already been introduced by Geoff earlier for gastric studies. This is an antibody drug conjugate. It is slightly different from dual anti-HER2 inhibition. This is a targeted chemotherapy. It uses HER2 as a homing beacon. This was a study that established its efficacy.

These were all patients that had been treated with multiple standard of care, treatment refractory, including prior anti-HER2 therapy. There were some of these patients that were actually treated with HER2.

You can see when you look at IHC 3+ and 2+, a bunch of patients responded. These are enviable waterfall plots as far as refractory colorectal cancer is concerned. Some of those responses last for a very long period of time.

When you look at the subgroup analysis, though, 1 of the most revealing findings is that the IHC 3+ population has a response rate of 57.5%. But the IHC 2+, despite the fact that it was FISH amplified, the response rate is only about 7.7%. This is a clear indication of the fact that it is the protein expression that is driving this efficacy rather than the amplification status, although, there is some correlation between amplification levels and the protein expression. So the higher the protein expression, the higher the gene amplification level. The lesson from this is that IHC 3+ are the ones that derive most benefit from this.

Another important thing is, despite the fact that some of these patients had had prior HER2, this drug was still active in that setting. This is not just a HER2 mechanism, but a true HER2 ADC. You can see response rates of 43% and 45% in that population, respectively.

[00:59:10]

DESTINY-CRC02: T-DXd for HER2-Amplified mCRC

This led to DESTINY-CRC02 study. This study was very critical for 2 points. Number 1, DESTINY-CRC01 study came at the background of MyPathway and tucatinib, and our understanding at that time was HER2 only works in RAS wild-type. That is what was enrolled on CRC01.

When we did the biomarker analysis for CRC01, we found that patients who were ctDNA positive for RAS mutations also benefited. That makes sense because this mechanism is not based on oncogenic addiction. It is based on protein expression. This study allowed RAS-mutant patients to be enrolled.

Furthermore, this study compared 2 important doses, 5.4 and 6.4, because the risk for ILD is lower with 5.4, as was demonstrated in breast cancer studies. Primary endpoint was confirmed objective response rate.

[01:00:00]

DESTINY-CRC02: Efficacy With T-DXd in HER2-Amplified mCRC

This is the efficacy analysis. You can see in the 5.4, response rates of 37.8%, very much similar to what was seen in CRC01. Median PFS and duration of response were 5.8 and 5.5 months. You can see that the 6.4 was comparable. So no decrease in the efficacy signal.

[01:00:24]

DESTINY-CRC02: Adjudicated Drug-Related ILD/Pneumonitis

More importantly, when you look at the side effects and toxicity in the 6.4 mg, specifically, the ILD occurred in 12.8% cases, whereas it only occurred in 8.4.

The critical point here is there was no grade 3 or more ILD events in the 5.4 mg/kg dose. That led to the approval of T-DXd in colorectal cancer at the dose of 5.4 mg.

[01:00:54]

Managing ILD/Pneumonitis With T-DXd

That brings us to a very important and critical point, which is management of ILD and pneumonitis. This drug is highly effective, but these side effects need to be recognized early. You have to have a very high degree of suspicion in the clinic. You have to screen your patients properly. Pulmonology needs to be get involved.

Broadly speaking, there are 2 types of ILDs that we see in the clinic. One is a grade 1 ILD. These are ILD patients that are asymptomatic and you just find them on a scan, which you are doing routinely. These patients have no symptoms. Only scan changes. Even if grade 1 occurs, you have to hold until those changes in the scans revert back so it becomes to grade 0.

If it is any grade 2 to 4, which is it is symptomatic, you permanently discontinue T-DXd. You never rechallenge with that drug.

Now, in grade 1 asymptomatic, there is data now that suggests you can actually rechallenge patients with T-DXd. If the changes resolve in less than 28 days, you can actually rechallenge at the same dose. But if it takes more than 28 days to get resolved, then you actually have to go down a dose level. Symptomatic ILD, discontinue T-DXd. Asymptomatic ILD, hold T-DXd. If it resolves really quickly, use the same dose. If it takes longer time, decrease the dose.

[01:02:24]

Trastuzumab Deruxtecan Can Overcome Resistance to Trastuzumab-Based Combinations

This can overcome resistance to prior anti-HER2 therapy, and we have already demonstrated that in the DESTINY-CRC02 study.

[01:02:33]

Posttest 2

I am going to go to the polling question. After progression on first-line chemotherapy and bevacizumab, a patient with metastatic colorectal cancer starts T-DXd. On the most recent scan, his right lung shows signs of asymptomatic ILD. In your current practice, how would you manage this patient with ILD?

1. Continue T-DXd at the same dose with close monitoring;
2. Continue T-DXd at reduced dose;
3. Hold T-DXd until resolved to grade 0 and then restart at the same dose if it resolves less than 28 days;
4. Hold T-DXd until it resolves and restart at a reduced dose, even if it resolves in less than 28 days; and
5. Permanently discontinue T-DXd.

Okay. It looks like pre-test, everyone either was discontinuing it or holding and reduced dose. And post-test, 100% people said restarted the starting dose because it resolved in less than 28. Great job, everyone. Thank you.

[01:03:59]

TRIUMPH: ctDNA-guided Pertuzumab + Trastuzumab for HER2-Amplified mCRC

These are again continuation of some of the data that further supports dual anti-HER2 therapy. This was a TRIUMPH study. It was done in Japan. What it did is they screened from tissue and they screened from ctDNA. You can find that the objective response rate is maintained at that 30% range.

[01:04:20]

Phase II Study: Zanidatamab + Chemotherapy in 1L HER2+ mCRC

Then the phase II study of zanidatamab, which was discussed with gastroesophageal, it has also been evaluated in combination with chemotherapy in first-line HER2-positive metastatic colorectal cancer. This was done in open phase II trial. These were all patients that were HER2 positive and they were given zanidatamab with modified FOLFOX-6 on every 28 days, or zanidatamab plus modified FOLFOX-6 and bevacizumab, with the primary endpoint being DLT.

[01:04:52]

Phase II Study: Zanidatamab + Chemotherapy in 1L HER2+ mCRC

You can see that this is a very small study, but it actually does show a promising impact of zanidatamab in combination with chemo and bevacizumab in patients like the response rates, as you can see, were close to 83% and nearly 100% when you have the triplet combination. This is a good development and we are looking forward to next studies regarding this.

[01:05:18]

Considerations for Initial Therapy of mCRC

As far as the initial therapy for colorectal cancer is concerned, there are multiple buckets of colorectal cancer. HER2 is 1 of those buckets. In the initial line, we do not yet have a HER2 approval but the current study with which MOUNTAINEER-03 is ongoing in second-line study, trastuzumab-pertuzumab or trastuzumab-tucatinib and trastuzumab deruxtecan are fair game.

[01:06:05]

HER2 in BTC

Dr Shroff: Last but not least, right in the middle of the GI tract, we are going to move into HER2 in biliary tract cancers.

[01:06:15] 

          Poll 10

We are going to start with a poll. How confident are you in your ability to select subsequent line treatment strategies for patients with biliary tract cancers following progression on frontline immunotherapy-based regimens?

1. Very confident;
2. Moderately confident;
3. Slightly confident; and
4. Not confident.

Please vote. All right. 50-50. Slightly confident and not confident. Well, hopefully we will shift that upwards to the very confident.

[01:07:01]

NCCN Recommendations for Molecular Testing and Incidence of Therapeutic Targets in Unresectable or Metastatic BTCs

Biliary tract cancers, as I already alluded, is a catch all phrase to include both intrahepatic and extrahepatic cholangiocarcinomas as well as gallbladder cancer. In total, they are called biliary tract cancers but obviously anatomically they are very different. Also what we have found over time, as we have learned more and more about the molecular complexities of these diseases is that they are also very different in terms of their biomarkers.

As you can see by this table, which honestly is not as comprehensive as we now know in terms of some of the potentially targetable alterations in biliary tract malignancies, there is actually some distinction in terms of the types of alterations that are seen based on the anatomic subtype.

I will obviously draw your attention to the HER2 overexpression and/or amplification, where as I mentioned, it is listed as being seen in gallbladder as well as intrahepatic and extrahepatic cholangiocarcinomas. It really is, I would say, predominantly more in extrahepatic cholangiocarcinomas and gallbladder cancers, again, in that 10% to 15% to up to 20% range.

[01:08:14]

HER2-Targeted Therapies in GI Cancers

We have seen this timeline, biliary tract cancers. While I love to call biliary tract cancers the model for precision oncology, we were a little bit later to the HER2 game than, say, obviously gastric cancer, but even colorectal. As has already been alluded, there is an approval for trastuzumab deruxtecan in a number of solid tumors based on the PanTumor data, and that includes a biliary tract cohort.

Then, of course, zanidatamab was also approved for biliary tract cancers. We actually had the HERIZON biliary tract data a little before the gastric data.

[01:08:51]

Phase II DESTINY-PanTumor02: T-DXd in HER2-Positive Solid Tumors

I will start with DESTINY. This was the phase II PanTumor02 that looked at T-DXd in HER2 positive solid tumors. We have already talked about this. This is the antibody drug conjugate. This was a multicohort phase II trial that included a group with locally advanced and metastatic biliary cancers.

Specifically, it was given at 5.4 mgs per kg and the primary endpoint was overall response rate.

[01:09:20]

DESTINY-PanTumor02: ORR and AEs With T-DXd (BTC Cohort)

When you look specifically at the biliary tract cohort, which was 41 patients, you will see that they looked at all comers and the confirmed overall response rate there was 22%. As Raghav just alluded, when you look specifically at what is admittedly a smaller number of patients, but the IHC 3+, that is really where we are seeing the bang for our buck with an overall response rate of over 56% and really nothing to speak about when it comes to IHC 2+.

Then again, the drug-related adverse events. Grade 3 was seen in a similar percentage, about 40% of patients. In terms of discontinuation, it was only 12.2% with T-DXd in the biliary cohort.

[01:10:08]

DESTINY-PanTumor02: PFS and OS in BTC Cohort

Again, the overall response rate was the primary endpoint, but importantly, median PFS and median OS. In totality, median PFS was 4.5 months with a median OS of 7 months. Again, really what we are seeing in those are the orange lines on both of these graphs is the IHC 3+ group, where median PFS was 7.4 months in a refractory patient population and 12.4 months in terms of median OS.

Just to level set, in the setting of second-line and beyond biliary tract cancers, our standard of care is really FOLFOX for all comers. There, we talk median PFS of 2-3 months, median OS of 5-ish, 5-6 months. These types of numbers in a subset of our patients is provocative and exciting.

[01:11:00]

HERIZON-BTC-01: Study Design 

Of course, that led to the tumor-agnostic approval for T-DXd. That is obviously an option for our patients with IHC 3+ positive biliary tract cancers. Then we also have the HERIZON-BTC-01 study. This was a single-arm, open-label, phase IIb study with the bispecific, zanidatamab, which, as Dr Ku already alluded, is really a biparatopic drug that specifically binds to 2 different sites on the HER2 receptor.

This was patients who had progressed on gemcitabine-containing chemotherapy, and it was 87 patients who were enrolled. There was 2 separate cohorts. But really it was cohort 1 that everybody was looking at, and that was a total of 80 patients planned that were IHC 2+ or 3+, that is how they defined HER2 positivity.

The primary endpoint again was confirmed ORR, with key secondary endpoints zanidatamab was given in the standard Day 1 and Day 15 dosing.

[01:12:04]

HERIZON-BTC-01: ORR by Subgroup by ICR (Cohort 1)

Now you can see this is the forest plot. All the way at the bottom, you will see the overall response rate. The overall response rate was just over 41% in the all comer patient population. Importantly, when you look specifically at that IHC 3+. As was already alluded to here earlier, the IHC 3+ is really the group that seems to be deriving the benefit. You can see that right here in this group right here as noted there. That is really the main thing that jumps out when you look at the subgroup analyses that were looked at.

Then I will also just point out again the numbers here. There was a nice mix of gallbladder cancer intrahepatic and extrahepatic cholangiocarcinomas with perhaps slightly more benefit in the gallbladder and extrahepatic, but small numbers to make those decisions.

[01:12:57]

Final Results From HERIZON-BTC-01: Zanidatamab in HER2+ BTC

With the median follow up of 33 months, this was the final data that we saw. Overall response rate stayed right at 41% in all patients, but in the IHC 3+ was 52%. Again, as was mentioned earlier, median duration of response in these patients was 14.9 months, which really in the world of biliary tract cancers was unheard of, especially even with our targets and some of our prior approvals that we had seen with FGFR-targeted therapies, IDH, etc..

Median PFS, specifically in the IHC 3+ patient population, as I mentioned, was 7.2 months.

[01:13:39]

Final Results From HERIZON-BTC-01: Zanidatamab in HER2+ BTC

Then overall survival again broken down by IHC. IHC 3+ median OS was 18.1 months. Again, I just want to level set that even with our frontline therapies right now, we talk median overall survival of about a year, 12 to 13 months. To see that type of median OS was really exciting, especially in that patient population.

As has already been mentioned, the main thing that we see with zanidatamab when we look at the safety data is diarrhea. You will see here that the majority of the patients had grade 1 and 2 diarrhea, but there was a smattering of that grade 3, albeit less than what was seen in the gastric data thus far.

As Geoff already mentioned, infusion-related reactions was an AE of interest specifically with zanidatamab. As you can see, again, these are relatively short-lived, easily managed, and are grade 1 and 2. I forgot how little there is to say about this in biliary cancer.

[01:14:39]

          Posttest 3

Post-test question. the following progression on a frontline ICI or immunotherapy-based regimen, which FDA-approved HER2-targeted therapy is appropriate for a patient with HER2-positive IHC 3+ metastatic biliary tract cancer?

1. Trastuzumab deruxtecan or tucatinib;
2. T-DXd or zanidatamab;
3. Tucatinib or zanidatamab; or
4. All 3, T-DXd, tucatinib or zanidatamab.

Fantastic. Okay, it looks like 80% said trastuzumab deruxtecan or zanidatamab, but there was 20% that said tucatinib. That might be alluding to a study with tucatinib and trastuzumab in biliary cancers. It was a multi-cohort study again that looked specifically at a biliary tract cohort. As Dr Raghav already mentioned, this is a tyrosine kinase inhibitor approached HER2 targeting.

There was some efficacy seen with a response rate, again, in the 30% range, much shorter duration of response. Then there was no further development of the drug, specifically in biliary tract cancers. So we have not seen an additional data there.

[01:16:16]

Emerging Trials in 1L HER2+ BTC

When we look specifically at what is coming or what is on the horizon for biliary tract cancers for HER2 targeting, the obvious question is, is can we use earlier introduction of HER2-targeting in biliary tract malignancies?

There is multiple studies ongoing, but the top 2 here are the big frontline studies that are being looked at right now. One is HERIZON. This is looking at moving zanidatamab earlier into frontline therapy with the standard of care, which is, gemcitabine platinum-based therapies with or without immunotherapy. This again is looking at the IHC 3+ patient population. This has a primary endpoint of PFS because there is crossover allowed.

DESTINY-BTC is the same concept. There, we are taking T-DXd moving it into the frontline space but combining it with rilvegostomig, which is a bispecific that basically includes the chemotherapy plus rilvegostomig plus T-DXd in HER2-positive, again, IHC 3+ patients.

Then I will just point out the SAFIR-ABC10, which is actually a really complex and really impressive effort across Europe that is looking at the concept of targeted maintenance therapies after standard of care. So upfront gemcitabine-cisplatin plus immunotherapy. Then based on your specific biomarker, you have various arms that you can move into maintenance. That does include some HER2 arms as well.

[01:17:53]

          Poll 11

Should we do experts in the hot seat?

Dr Raghav: There is actually a bunch of questions. Maybe we can take the question.

Dr Shroff: Okay.

Dr Ku: Yes. Time for questions.

Dr Shroff: I am handing that off to you to moderate.

Dr Ku: Yes, I can moderate the questions.

[01:18:02]

Q&A

Dr Ku: Actually, this is a great question to follow on from your talk. How do you choose between T-DXd and zanidatamab?

Dr Shroff: That is a great question. Nobody knows is the short answer.

Dr Ku: All right. Good night, everyone.

Dr Shroff: First of all, we had T-DXd available to us with approval before we had zanidatamab. Zanidatamab’s approval just came in in the latter half of that year. I would say that people gravitated towards using that first because it was available first. At this point in time, it has a lot to do with just specific patients and the toxicity profiles, risks and patient preference.

In the gastric study, they do diarrhea prophylaxis with the zanidatamab and there is a lot of concerns, especially in biliary malignancies in terms of just GI dysmotility and things like that. Those are conversations that we have with our patients. But the truth of the matter is, is we do not know.

Dr Ku: I mean, in your clinic, I am sure you have tried both sequences. Anecdotally, anything rise to the top in terms of differential activity toxicity?

Dr Shroff: I will say that the zanidatamab efficacy that you see in the trial. I have seen really nice durable responses with zanidatamab. The obvious other question then is, after one, can you give the other? I have done it is the honest truth. I have not seen a whole lot of wonderful, exciting activity giving zanidatamab after T-DXd. I had 1 patient that had a really nice response after zanidatamab to T-DXd.

Dr Ku: I think this is a question for the both of you. What are the data regarding HER2 loss after being exposed to anti-HER2 therapy?

Dr Raghav: Yes. I think that after T-DXd, there is definitely some evidence of HER2 loss that happens. I do not know whether that happens purely because it is HER2 T-DXd, or is it because most of these patients have had HER2 therapy for a longer period of time.

In colorectal cancer, HER2 expression is more homogeneous, so the loss is probably less. There is also a concern that in most cases when we talk about this loss, we are talking about like ctDNA and the sensitivity goes down because there have been cases where initially they had lost it and then they have gained it back. So hard to say.

The lesson is always test it before you are going for that next HER2 treatment for you.

Dr Ku: Rachna, for you?

Dr Shroff: Yes, similar. There is not clear data that there is change. Because as I mentioned earlier that we do not necessarily have enough tissue to really do a whole lot of sequential testing, it is really more ctDNA based, and so it is hard to know for certain, but we do not think that there is a significant change.

Dr Ku: Got it. Yes. Actually 1 question that came through, for me was whether there is evidence of HER2 loss in esophagogastric when you are receiving non-HER2-based therapies. There is not a lot of data, but I would be inclined to say the answer is no. Having said that, that would be a comparatively rare situation. Maybe the situation would arise if someone has a localized tumor for which anti-HER2 therapy is not standard, and then they then develop metastatic disease. The question is whether the tumor retains HER2 positivity?

Again, anecdotally, I would say yes. HER2 is really lost as a mechanism of resistance selection pressure from anti-HER2 therapy.

We have time for 1 more question. It is, what role will HERIZON-GEA regimens play in the setting of patients that have already progressed on trastuzumab?

I would say that there is no role. Assuming someone receives KEYNOTE-811 at this point and they progress, the standard treatment in that setting is trastuzumab deruxtecan. Now, having said that, in the original phase I study with zanidatamab, it was treatment refractory HER2 positive cancers. In the HER2-positive gastric cancers, that had received prior trastuzumab, the monotherapy activity for zanidatamab in that population was still in about the 35% range, but certainly that is not an FDA-approved indication. It is not an NCCN compendium listing.

I would say that the HERIZON-GEA regimen is meant as a first-line regimen. Therefore, certainly it has no established role in the second-line setting.