Introduction

DR LOVE: Good evening, everyone. I’m Neil Love from Research To Practice, and welcome to Beyond the Guidelines, as tonight we talk about the management of HER2-low breast cancer. We have an outstanding faculty, and later on in the program we’re going to be looking at a survey of the faculty, as well as some other investigators, as we go through this evening. We will be talking about indications and use of medications. We’ll be focused on risk/benefit profiles of the therapies that we’re talking about tonight.

We’ll be in San Deigo tomorrow morning bright and early starting with 4 symposia there, so for those of you who are online, or if you’re here in San Antonio, and you want to check out some ASH content, or hematologic content, we’re starting out early in the morning with non-Hodgkin lymphoma, moving on to CLL, and by midnight on the eastern coast we’ll be finished with multiple myeloma. So check it out. So many things going on today in oncology.

And not to mention we’re reawakening our weekend-long symposia in our hometown of Miami next March. So for general medical oncologists, we’ll have more than 25 investigators in different tumor types there to talk about taking care of patients.

For the attendees here in the room on your iPads all of these slides of the faculty are there. You can leaf through them. And also there’s a survey for you to take. If you take that you’ll get a lot more out of this tonight. If you have any questions or cases you can also submit them through the iPad, and you can also complete the evaluation. And all of these functions are also available for our online audience. Hello to everybody. We’ve had a lot of people online this week. This is actually the third of 3 programs we’re doing here in this room. So all of those functions are available to those of you who are online tonight.

We are video recording all 7 of these symposia this week at San Antonio and ASH. We’ll let you know by email when they’re posted for your colleagues to check them out, and for you as well.

But today we’re talking about HER2-low breast cancer, something that kind of really didn’t exist a couple of years ago. Dr Modi and others brought this data to our attention, and it’s really been fascinating.

So as we do with a lot of our programs we conducted an interview of investigators. We’re going to show you the results of that. We asked them about their usual treatment practices. It’s our way of kind of getting beyond the guidelines. The guidelines will present you with a number of evidence-based options. We like to just go to people and say, well, what do you usually do in this situation. If everybody usually does the same thing we call it a consensus. If there are 3 or 4 options, well, a little bit more to talk about.

The other feature that we’re going to bring out tonight, and we’ve been doing all this week, is bringing some additional faculty to your attention. I worked with these 6 investigators over the last few weeks. I met with each one of them individually for about an hour and a half, just chatted with them about breast cancer and basically said what do you think we ought to talk about in these 3 meetings, asked them about their treatment patterns, and basically tried to get some rounds content here for our faculty to react to. And we’ll be showing those.

All these videos that we’re showing you tonight are less than 90 seconds. You’ve heard of TikTok? This is TikTOnc.

Optimal Approaches to HER2 Testing for the Identification of HER2-Low Breast Cancer — Dr Carey

DR LOVE: So here’s where we’re heading. Each of the faculty will be giving a more traditional presentation, but prior to that we’ll go through some clinical issues. So we’re going to start out in a second with Dr Carey talking about testing.

So Lisa, we actually asked the faculty — we have numbers from the literature. We asked people, now that you’ve been out there looking for HER2 low, what have you seen in both in hormone receptor-positive and hormone receptor-negative, and we put the estimates. Of course HER2 low’s a lot more common in the hormone-receptor positive, and Lisa will be reviewing. So much to talk about in terms of HER2 testing.

And then the other thing that was kind of interesting, Lisa, is the question of whether you see benefits outside of breast cancer. We’re starting to hear about the use of — we’ve been hearing about anti-HER therapy in other solid tumors for quite a while. And of course now they’re out there, not only do they have approved therapies in GI cancers and lung cancer, and GYN is coming along fast, looking for HER2 low. Any thoughts about whether any response has been seen? I think I saw in biliary? I’m not even sure how much people are looking at FISH in these things, too. Any thoughts about that?

DR CAREY: I mean, I think my thoughts are it’s the Wild West right now. I think there’s going to be a coming to grips to how we actually examine HER2 and particularly the ADCs that are HER2 targeted that’s going to change, and I think that’s going to be true in breast cancer. It’s certainly going to be true in the other tumors as well.

DR LOVE: Aditya, any thoughts about this? I should mention, too, I found out in the faculty room, actually you may have heard that last night we had Sara Hurvitz, who was at UCLA, and then Tom Lynch from the Hutch, a lung cancer doc, stole Sara from UCLA, but now Dennis Slaman grabbed Aditya from MGH, and you’re heading out to UCLA, right? You’re going to be a Bruin.

DR BARDIA: Wild, Wild West.

DR LOVE: Wild West, right? And you’ve got your own HER2 team because we work with Zev Wainberg in GI cancer, very into HER2. Randy Hecht is there, colon, Ed Garon in lung. So you’ve got a great HER2 team to join. Dr Kenichi.

Anyhow, getting back to HER2 low outside of breast cancer. Any thoughts? Any thoughts about why these patients are actually responding? I mean at a basic level one of the issues here is like what’s going on with HER2 low in a way that you can explain to us. What’s your vision about why it works?

DR BARDIA: Yeah. I think part of the problem is the terminology of HER2 low. Essentially what it means is that it’s a HER2-expressing tumor, and we know that trastuzumab deruxtecan, when it binds to HER2, it releases deruxtecan, which has a bystander effect. So as long as the tumor has some expression of HER2 because of the bystander effect it can exert efficacy, and that’s not restricted to breast cancer.

We saw from PanTumor01 that trastuzumab deruxtecan had efficacy in other HER2-expressing tumors as well. You mentioned biliary cancers, other rare cancers as well. And this is really important for rare cancers because it’ll be impossible to do randomized clinical trials for rare cancers. If you have good single-arm study data I think that would be sufficient for the use of these agents.

DR LOVE: And Peter, that ties into the question. We’re seeing a number of pan tumor approvals, BRAF, obviously IOs for MSI-high disease, and we asked in this survey — we put some of the verbatim answers in the survey in there, as well, for you to read. Also you can just leaf through the survey. I think it’s very instructive.

But we asked them do you think that T-DXd should receive a tumor agnostic approval, and actually more than half of the faculty says yes. Any thoughts about that, Peter, and what it takes to be able to try something or use something outside, in a different tumor?

PROF SCHMID: Certainly it’s an important questions. It’s tempting to say it should be tumor agnostic. On the other hand you have to consider the efficacy in each tumor type is not just dependent on the expression but also in how the payload works, and that can be dependent on how those tumors are pretreated. And there can obviously be overlap for some of the pretreatments.

So I do think it is important to establish efficacy in different tumor types. Do you need to go through big Phase III trials in every tumor type? That’s a different discussion. But I think you should show at least efficacy that’s similar to other tumor types observed.

DR LOVE: So Shanu, when you presented the HER2-low data, Dr LoRusso brought up something we’d already been hearing from the lung cancer people in terms of HER2-mutant disease, where actually in lung cancer right now that’s where it is. T-DXd is actually indicated use. Some of the investigators want to use it first line, very effective in lung cancer, but there it’s HER2 mutant. I know HER2 mutant is very unusual in breast cancer, but do you think that that’s a reason to use T-DXd, Shanu?

DR MODI: I mean there is a study ongoing right now that is looking at, as you were saying, in an agnostic way, enrolling patients with HER2 mutations to T-DXd. I mean it’s really exciting to see the lung cancer data, and I suspect it’s going to work in other settings as well. We’ve seen that TKIs can certainly work in the HER2-mutant breast cancer population. So I’m optimistic. It’d be great to have something for that population as well.

DR LOVE: Yeah. I’ve seen data with neratinib. Any thoughts, Shanu, whether or not a TKI or T-DXd, theoretically, would have greater activity in these HER2-mutant patients?

DR MODI: I mean I think what we know about these HER2 mutants is that there is rapid turnover of the receptor. So I think there is real potential there. Maybe a combination of both is really what we need to do.

DR LOVE: All right. Well let’s go to our auxiliary faculty, and Lisa I’d like to hear your thoughts about what Dr Brufsky and Meisel have to say.

DR BRUFSKY: Generally when someone comes to me, I’ll look at the IHC first, and as we know from DB-04, it doesn’t really matter when that IHC was done. It could be in a primary. It could be in anything.

And I think that we get a lot of milage out of asking the pathologists if it’s HER2 0, for example. Take another look at the slides, potentially recut a block and give us a few more slides and re-stain them. It doesn’t matter if it’s the primary or a met. I think at the end of the day, it’s going to be any faint staining on any slide is probably going to be enough for me to try trastuzumab deruxtecan in that setting.

DR LOVE: Have you had patients that you really believe were HER2 0 and therefore did not use T-DXd?

DR BRUFSKY: Very, very uncommonly. I really, really push as hard as I can to get T-DXd.

DR MEISEL: For me, if anyone has any history of HER2 low in their pathology reports, even dating back to their original tumor, I feel like it’s probably worth a try of T-DXd. My experience with the drug has been that when it works it does tend to work fairly quickly, and it’s not usually subtle whether it’s working or not. So you can always backtrack and go off of it if the patient doesn’t respond.

DR LOVE: So Lisa, any thoughts?

DR CAREY: I totally agree with them, and in truth the existing data suggests that when pathologists are trained on this they actually — the biggest shift in the scoring is from 0 to low. So I think we’ll be seeing more of them.

DR LOVE: I guess you’re going to get into this in more granularity in your talk, but I mean have you had patients who needed a response, ran out of therapies, and were HER2 0, and you did not use T-DXd?

DR CAREY: I agree with Adam. I try everything to get HER2 low out of something, and then I try it.

DR LOVE: All right. Well here are a couple more comments, and Peter, I’d like to hear what you think about what Drs Tarantino and Winer have to say.

DR TARANTINO: It’s very rare to have a patient with all of the biopsies being HER2 0. Not impossible. We do see some patients like that, especially if they have triple-negative disease.

Sometimes it’s really basal, non-HER2 expressing, whereas most of the hormone receptor-positive cases have at least 1 biopsy that is HER2-low.

But in general I do feel that the data are very strong and will lead to agnostic approval.

DR WINER: My colleague at Yale, David Rimm, has been working on a more quantitative approach to measuring HER2 low.

As you know, at this point in time, it’s up to the pathologist in terms of just determining what’s 1+ and what’s 2+. And in many cases, pathologists are asked to go back and look at specimens. And it feels very awkward to them. So I think we do need to do better in terms of coming up with a better way of defining what HER2 low is unless it turns out, of course, and this could be the case, that drugs like trastuzumab deruxtecan work in HER2 any. It may be that you don’t need 1+ or 2+ and that HER2 0 is okay.

I think that at the moment, if somebody has HER2 0 disease repeatedly and there’s never been any evidence of any HER2 staining, #1 the drug isn’t approved in that setting, and I probably wouldn’t use it. Do I think that it’s crazy to do that? No, I don’t think it’s crazy.

DR LOVE: So Peter, you’re great at explaining biology to people like me who have a hard time understanding it, but we have ADCs targeted, like targeting TROP2, except we don’t measure TROP2, and then we have some that are targeted like HER2 except we’re just trying to find anything to justify using it. Any thoughts about where this is going to land?

PROF SCHMID: So I think there are a few things to consider. First of all, with this new generation of ADCs you only need very few antigens on a few cells to make that drug work because it will not just tackle 1 cell but also the surrounding cells, the bystander effect. The second thing to consider is HER2 0 doesn’t mean there’s zero HER2 expression. It just means it’s very, very low expression, and there are some tumors that have absolutely no HER2, but there are many tumors that have very low HER2 expression.

The third thing to consider is also the pattern how you find positivity. You can have a lot of positive cells clustered together and then vast areas of the cancer negative, or you can have the odd positive cell interspersed in the cancer. If you have some activity and some distribution of positive cells I think the treatment will work. We will learn a lot. I think there’s a group of HER2 0 patients that will benefit from T-DXd, whether it’s all patients I think we will learn.

DR LOVE: Great. All right, Lisa. Let’s take a look at some data.

DR CAREY: All right. So I get to talk about defining HER2 low. So let’s first by stipulating that HER2 testing has not been a static thing. It’s been a lot of changes over the years in how we interpret immunostains and ISH. But in the end of the day the purpose of everything we’ve done up until now has been to optimize the use of anti-HER2 drugs, specifically things like trastuzumab inhibitory antibodies. But that all ended with DB-04 when now the purpose is twofold. One is you still have to optimize anti-HER2 drugs, but now we have to identify HER2-low.

So let’s start with immunostains because everything about this is immunostains at the moment. What you see here is the classic algorithm for HER2 with the positive being 3+, which is really defined by how much staining and how deeply it stains. Being 3+ and 2+ with ISH, so gene copy number amplification being positive and everything else being negative, but now we have the low category, which is things with some, that 1 and 2+ without ISH.

So again, the focus for the first 20 years of this story was about defining where inhibitory antibodies would work, so it’s HER2 positive versus everybody else. But then came the ADCs, and now we have to define HER2 low because they clearly work in this middle area.

So HER2 low, of course, is 1+ and 2+ that’s negative for ISH, so now it’s the next register down. And this was not a new concept. It was defined in NSABP B-47, where they tried to see whether trastuzumab in the adjuvant setting could work in HER2 low. It did not. But that’s what they used in the ADC trials. And it’s essentially below the threshold where I’m calling them naked antibodies would work.

Now of course now we have to take these definitions and reconsider them. So I’m showing you here the 2023 ESMO Clinical Practice Guidelines, which took the traditional HER2 positive, HER2 negative and then moved it into HER2 positive, HER2 low. And then within HER2 0 they proposed another categorization of HER2 null, which is basically HER2 really 0, and HER2 ultralow, which is not completely 0 but not quite to 1+. And that actually is the definition that’s included in DESTINY-Breast06, so we’ll find out more. At the moment, to be frank, it’s not clinically validated, so we don’t quite know whether that’s going to hold up.

And this gives me an opportunity to say okay, don’t forget, assay development is complicated. You can’t make it up in your garage. You have to know that you have analytic validity of your assay, you have to know that there’s clinical validity that the assay actually differentiates tumors into real things, and then you have to decide whether it’s clinically useful, meaning you use the assay, you make better treatment decisions.

So our pathology colleagues have not bought into this whole thing, and this is a direct quote from them. And you can see here this guideline update, which is the 2023 HER2 update, does not support the use of HER2 low because DESTINY-Breast04 didn’t include 0. We don’t know if 1+ or 2+ versus 0 actually matters for trastuzumab deruxtecan response because it wasn’t tested in the FDA expansion just because of a clinical trial not because of a biologic criterion.

So a couple of comments before I jump into that. HER2 testing was designed, just to say the obvious again, to differentiate HER2 driven, HER2 addicted, HER2 whatever, what we have traditionally called HER2 positive, from everything else. It really wasn’t designed for what we’re asking it to do right now.

HER2 low was not ever defined as a biologically distinct entity. It was a byproduct of the trials, a thoughtful one though because it’s plausible that some HER2 expression would be necessary for a drug that targets HER2 as the antibody part, but how much is not clear.

So now let’s see if we can differentiate this a little bit. So clinical characteristics of HER2 low versus 0. As was noted, ER positive and ER negative, that’s a big one. So most ER positives that are traditionally HER2 negative are in fact ER low. Most ER negatives are not HER2 low. It’s about a two third/one third split.

Now beyond that, if you look on the right at the National Cancer Database, when you get past that there’s pretty subtle stuff, and it’s not very compelling different. A little bit of age, a little bit of grade, but pretty similar clinical characteristics.

Prognostically also very similar. Now just to be clear, this is conventional therapy prognosis. This is not specific to antibody-drug conjugates targeting HER2.

What about molecular subtyping? Are there molecular differences? Well, not a lot. So luminal A a little bit more represented in HER2 low versus 0, but only a few percentage points. Similarly, basal-like a little bit more in 0 than low. And almost surprisingly not much HER2 enriched in either one of them. But pretty subtle stuff.

And this may be part of the problem. The testing itself is very hard to do in the lower levels of HER2 expression. And what’s on the left is HER2 low by a bunch of high-volume Danish pathology departments. And I’ll point out if you look on the right-hand side of all the little spider lines, on the top is 1 department, green line, on the bottom is another department, and all of the other ones in the middle, and there’s a huge range in the proportion of tumors that they’re calling HER2 low.

On the right you see 18 board certified pathologists. And the right plot is HER2 3+ versus not 3+, and they’re pretty concordant in that. On the left is HER2 0 versus not 0, and there’s a lot less concordance. So there’s a lot of intraobserver variability in calling HER2 low. It’s not the same as HER2 3+. And there’s reasons for this. If you get into the lower areas, analytic variations and fixation, antigen retrieval, histologic and spacial heterogeneity, et cetera, all start to play a bigger role.

And because of that the ESMO expert consensus statement basically said given these considerations HER2 low disease should not be considered a distinct molecular entity by a heterogeneous group of tumors with biology mostly driven by the hormone receptor. And you can see it’s a 94% concordance.

Can you get better? Yes. The short answer is you can. About 800 tumors from metastatic patients that were rescored after formal retraining in low-level immunostaining using the usual ASCO/CAP Guidelines. You can see it had about a 67% HER2-low proportion. It didn’t matter what antibody they used. And there was about 80% concordance between the original score and the rescore, so they definitely changed. And in particular, as I alluded earlier, it was particularly from HER2 0 to HER2 low. In a separate analysis 74 pathologists who had 4 hours of extra training the HER2 0 concordance amongst them went up by more than 10%.

Now there’s new approaches that are coming. So to be honest, I suspect we’re not going to be using ASCO/CAP 2018 Guidelines for HER2 testing in the future. I think we’re going to have better systems. Eric Winer alluded to AQUA, which is David Rimm’s quantitative fluorescence approach, which is in fact — we know improves discrimination in the lower HER2-expressing tumors. There has also been a bunch of analyses using machine learning, which is good, attached to image analysis that reduces interobserver variability.

So to summarize, there’s little evidence that HER2 low actually represents a distinct biological entity, but it is clinically actionable, and any HER2-low test counts from a clinician standpoint. Conventional immunostains, the definition is improving, but it may be suboptimal for both technical and biologic reasons. And newer methodologies may, and I suspect they will, provide better clinical validity and utility, and the threshold is currently unknown.

Available Data with and Current Role of HER2-Targeted Therapy for HER2-Low Disease — Dr Modi

DR LOVE: Thanks a lot, Lisa. We’re going to move on now and kind of get into clinical issues involved in managing these patients, particularly as it relates to the use of the antibody-drug conjugate T-DXd.

So back to our survey, Dr Modi. So first question relates to the ER positive, the more common variation of HER2 low, and you see that the investigators are pretty centered around trying 1 line of chemotherapy, I would think most likely capecitabine first. When we ask the same question related to hormone receptor negative, they’re kind of split between after 1 line of chemo or after 2. Any thoughts in terms of line of therapy based on ER?

DR MODI: I mean I think what you’re seeing is people are using this drug based on how it’s approved, actually, really after 1 line of chemotherapy. And I think that’s entirely appropriate, frankly. I mean for our hormone receptor-positive patients, I mean I think we really want to try to use our targeted therapies, our best, well-tolerated targeted therapies, and so for these patients that’s endocrine therapy. And if there are those tumors that we can identify that are less sensitive to endocrine therapy it would be great to move quickly to chemotherapy-type regimens and antibody-drug conjugates. I think you’re going to see results from the DESTINY-06, maybe to support that kind of a strategy.

So I think what you’re seeing is the audience following the guidelines, following how this drug was studied in clinical trials. And overall I think it’s a rational approach to using ADC therapy today.

DR LOVE: So being more specific, Aditya, we said okay, what comes first. And of course, I think actually now that I think about it, we were doing an ASCO satellite about an hour after you presented the data. So they were like — had just seen it. Everybody walked in the room, but they said the same thing then that they’re saying right now, which is, Aditya — and I don’t know whether you agree with this, but by the way we do it, it looks like there’s a consensus in terms of the key question of which antibody-drug conjugate comes first, T-DXd or saci, with ER positive T-DXd, ER negative people leaning more towards saci. Is that kind of the way you think it through?

DR BARDIA: Yeah. I fully agree because the DESTINY-Breast04 trial showed a very impressive improvement in progression-free survival and overall survival with T-DXd, and 90% of the patients enrolled in DESTINY-Breast04 had ER-positive disease, and that was the primary endpoint as well. And given the magnitude of benefit we’ve seen with T-DXd most physicians would use T-DXd preferentially before sacituzumab govitecan for ER-positive disease.

For ER-negative disease we just don’t have the same level of evidence as compared to the ASCENT trial, which looked at sacituzumab govitecan, because DESTINY-Breast04 only had 10% of patients who had ER-negative disease. And maybe that’s why you’re seeing physicians choose sacituzumab govitecan before T-DXd in that setting.

DR LOVE: And I just showed a bunch of comments that were in the survey. You might want to read through these. We have a lot of verbatim comments talking about their experiences with HER2 low and responses. A lot of people, including some of our faculty on video, talking about very impressive responses.

So Peter, I’d like you to respond to Dr Sharma’s thoughts about using T-DXd in this situation.

DR SHARMA: If you look at cross-trial comparisons the T-DXd data looks a little more robust. So if it’s a patient who has HER2-low disease I’d be more inclined to prescribe T-DXd assuming they don’t have any lung issues or ILD or other things that I’m worried about, and then sequence the other drugs after that. If it’s a patient who does not have HER2-low disease, then dato-DXd and sacituzumab are both viable options for hormone-positive metastatic breast cancer.

And looking at cross-trial comparisons between TROPiCS and the dato-DXd TROPION-Breast01 trial, the hazard ratio for improvement in PFS is similar between the 2 trials. We haven’t seen the overall survival data from the TROPION, but we’ve seen it from TROPiCS. So again, we’ll have choices, and it will depend on the toxicity of the drugs. And ultimately our institutions’ pathways and insurance carriers might start dictating what we use in what sequence.

For triple-negative I still favor saci just because very small numbers in the DB-O4 with triple-negative, HER2-low disease, and the saci ASCENT trial was a large, randomized study of only triple-negative patients.

For hormone positive, HER2 low I sequence T-DXd before saci.

DR LOVE: So Peter, any thoughts? Again, kind of putting aside access or reimbursement, but just based on what you think is best for the patient. Any comments about how you sequence T-DXd?

PROF SCHMID: I think that was an excellent summary. Ultimately our strategy at the moment is reflective of the amount of data we have not necessarily of the efficacy of the drug. We have more data for sacituzumab in triple-negative breast cancer than for T-DXd. That’s why we would choose that drug in sequencing first before we move to T-DXd. This can reverse if more data with T-DXd were coming out.

In ER-positive disease it’s the other way around, and also the T-DXd study was in an earlier line compared to sacituzumab, and that’s why most colleagues would prefer T-DXd as the early option and the sacituzumab as the second option.

But ultimately the optimal sequencing we have to learn, work out over the coming years. And one of the things I think we’re not paying enough attention to at the moment is the payload because no one would give paclitaxel follow up by docetaxel followed by nab paclitaxel, and to some degree this is what we’re trying to do with some of the ADCs. So we need to focus on getting other ADCs in with different payloads to overcome some of the resistance.

DR LOVE: Yeah. A number of faculty this week have kind of brought up that issue, and I guess a lot of uncertainty about what to do about it.

Lisa, I’m curious what your thoughts are about Dr Brusfsky’s comment.

DR BRUFSKY: There’s a lot more in ASCENT than DB-04. And so for that reason I tend to use saci first in triple negative and then trastuzumab deruxtecan second.

Now what’s interesting, and I’m curious, I don’t know what my colleagues think about this. Some of my colleagues now have said we need to interrupt the antibodies. We can’t give one after the other. We need to do chemo in between. And I’m trying to understand why you need some sort of rest between ADCs. I don’t know the answer behind it. I’m curious to know that and there’s probably something that I just don’t know about it yet.

DR BRUFSKY: I’ve had a woman, I put her through everything and then trastuzumab deruxtecan. And I treated her with it, and she had a complete response in her liver. So you do see responses like this, and I’ve seen many. That’s why I think it’s just such an interesting drug because you salvage people who have just been through everything and in the past we would just kind of recycle therapy. They’ve been through everything we’d had, like seventh, eighth-line therapy and they respond to this.

I’ve had some very long responders and some CRs.

DR LOVE: So Lisa, any thoughts? And maybe you can talk a little bit about your vision about mechanisms of resistance to both chemo and ADCs, what we know about it.

DR CAREY: Well, you’ve just hit the nail on the head, right? The issue here is you can, and there’s been not a lot of data, but some data already with sacituzumab, that there are acquired resistance mechanisms on the basis of TROP2 mutations and also on the basis of TOPO1 mutations. And so you could become resistant to the target mechanism, or you could become resistant to the payload mechanism, and so it gets a lot more complicated.

But I think what Adam alluded to is a really interesting thing, and we saw this, I don’t know, in the poster session today, several of the investigators from SWOG and from Alliance, did a survey of how are people sequencing ADCs. And it was really interesting. The truth is that intervening with chemo cropped up a few times in some of the case-based things, which I was a little surprised at that. I didn’t know where that came from, and he’s obviously hearing that too.

I’m not sure it makes sense to me. I think I’m not sure that if a patient — if I was worried about a patient, and they had acquired resistance to 1 TROP2-directed ADC, I don’t know that I’d reach for another TROP2-directed ADC, right, immediately afterward. I might try to intervene with something else. But the difference between saci and T-DXd are you’ve got 2 different targets, and you do have a shared TOPO1 mechanism, but they are different drugs.

DR LOVE: So Aditya, I’m curious what your thoughts are about Dr Winer’s comment, again, in terms of sequencing. It was interesting, when I sat down with him he had just run this conference in Lisbon, kind of a consensus thing. Really cool. I love the way they plot their answers, incidentally. And I’d be curious what you think about Eric’s thought.

DR WINER: I think in that setting, at this moment in time, if a patient has HER2-low disease, I would then reach for trastuzumab deruxtecan. And I would give that prior to giving sacituzumab or anything else. And that was actually the preference at the recent ABC meeting when we ran a consensus conference. And this was in Lisbon a week ago. And the question was asked in an ER-positive patient who has HER2-low expression, what’s your next drug after capecitabine? And the answer in well over 90% of respondents was T-DXd which was a different answer than the answer in a triple-negative patient who had HER2-low expression, but where given the numbers of patients in the various trials and the efficacy of sacituzumab there, the choice was sacituzumab first.

DR LOVE: So Aditya, any thoughts?

DR BARDIA: I fully agree. I was not in Lisbon, but I agree with the ABC consensus that we should use T-DXd first before SG for patients with ER-positive, HER2-low disease and the oppositive for triple negative based on the level of evidence.

DR LOVE: All right. Well let’s take a look at some data, and Dr Modi’s going to go through some of her work, as well as some other work that has been done looking at this.

DR MODI: So I think as Lisa really nicely already described, we know historically we’ve classified breast cancer as either HER2 positive or HER2 negative, but we know within that HER2-negative population there are these tumors that have low levels of HER2, and we refer to them as HER2-low breast cancer even though ASCO/CAP doesn’t acknowledge that. And we’re trying to understand more, obviously, and maybe define these cancers better. But what’s interesting is even our HER2 targeted therapies work extremely well for HER2-positive breast cancers, but they haven’t — our historically available drugs haven’t worked well against this low level of HER2.

And as Lisa already referred to, the NSABP B-47 study is probably the most definitive, I think, example of that. So this was a large prospective, randomized trial. Over 3,000 patients with HER2 low, early-stage breast cancer were randomized to either chemotherapy or chemotherapy plus 1 year of trastuzumab. And as you can see here in the efficacy curve there was absolutely no benefit to adding 1 year of trastuzumab to these patients who had HER2-low breast cancer.

And we’ve seen similar negative results for pertuzumab, this time in HER2-low metastatic breast cancer, where the response rate was less than 5%. And similarly even with T-DM1, of course an antibody-drug conjugate, which is, again, using a different mechanism, using the HER2 as an anchor to bring a payload, and yet even in this case T-DM1 works great in HER2-positive breast cancer but very little impact in HER2-normal or HER2-low breast cancer patients. So we’ve never had HER2-targeted therapy to offer these patients with tumors that have low levels of HER2 until this new generation of HER2 antibody-drug conjugates.

And the field is really being led by trastuzumab deruxtecan, or T-DXd for short. So we call this a next-generation ADC. It has some unique, I think, differences, advantages, compared to the first-generation T-DM1 ADC. So they both have trastuzumab as a backbone, but the payload is different. The T-DXd has a TOPO1 inhibitor, which is a kind of chemo we don’t often use in breast cancer. It also has 8 chemo molecules linked to each antibody, so that’s twice what we see with T-DM1. So it delivers a lot more chemotherapy to the cancer cells.

And then most importantly it has that cleavable linker, and so when it releases chemo, the chemo payload, that payload remains membrane permeable, and now it can produce what’s called the bystander effect.

And to show you how that works. So here the ADC binds to its target, HER2. The whole complex gets internalized into the cell, and the chemo is released, and it kills the HER2-positive cancer cell. That’s where T-DM1 would stop working, but in the case of T-DXd that chemotherapy can now pass through the cell membrane, enter the tumor microenvironment, and neighboring cells regardless of their HER2 expression. And that’s called the bystander effect, and it allows T-DXd to be active in tumors with variable levels of HER2 expression, including low levels.

And so based on that unique mechanism of action we did a proof of pilot Phase I study of T-DXd in patients with HER2-low metastatic breast cancer and saw a pretty exciting signal of activity, and that’s what eventually launched the DESTINY-Breast04 trial, which is a global Phase III study of T-DXd versus standard chemotherapy for patients with HER2-low metastatic breast cancer. And we presented the primary analysis in 2022. That was with 18 months of follow up. In this past ESMO we presented an update on survival from this study with now 32 months of follow-up.

So just to remind you, this was a study for patients who had centrally confirmed HER2-low metastatic breast cancer. They had 1 or 2 lines of prior chemo. And for the hormone-positive population they had to have prior endocrine therapy and be considered to have endocrine-refractory disease. Randomized 2:1 to T-DXd or standard chemo of physician’s choice, and the primary endpoint was to look at progression-free survival, first in the hormone-positive patients, then for all patients on trial, and similarly for overall survival.

As Aditya mentioned, 90% were hormone-positive breast cancer patients on this study, 10, 11% were hormone receptor negative. The hormone positive are in the left 2 columns. All patients are in the right 2 columns there. And just to briefly say, the groups are really well balanced. On average patients had 3 lines of systemic therapy. 60% had 1 line of chemo, 40% had 2 prior lines of chemo. For the hormone-positive patients they had an average of 2 lines of prior endocrine therapy, and 70% had prior CDK4/6 inhibitors. And overall the patients on the trial, 60%, had HER2 1+ cancers, and 40% had HER2 2+ cancers.

And these are the primary efficacy endpoint curves on the left. For the hormone-positive population you can see the hazard ratio for PFS was 0.51, a clear advantage for T-DXd. We saw the median PFS improve from approximately 5 months to 10 months. And the results were nearly identical when we looked at all patients on the trial, including the hormone receptor-negative patients.

So in subgroup analysis again you can see very consistent advantage for T-DXd across all subgroups, and I will point out specifically this was true regardless whether patients had CDK4/6 inhibitor therapy or not, whether they had an IHC 1+ tumor or a 2+ tumor, and whether they had 1 or 2 lines of chemotherapy.

So at this past ESMO we presented updated survival. This is now 32 months of follow up, and it was great to see the results were very consistent with the primary analysis. So on the left, for the hormone-positive population, again this is overall survival now, the hazard ratio was 0.69. And overall the median survival improves from roughly 17, 18 months up to 2 years with T-DXd. And on the right are all patients on the trial. Again, very similar. There’s about an absolute gain in overall survival of 6 months.

We also updated the hormone receptor-negative cohort, that triple-negative cohort that we’ve been talking about. And this was a small group of patients. This was an exploratory analysis. It was roughly 58 patients. And you can see here, although this isn’t a statistically rigorous analysis, I think you can see the results are fully aligned with what we saw for the overall study population, really striking efficacy in favor of T-DXd. The OS hazard was 0.58. The PFS hazard by investigator assessment was 0.3. So these are really clinically meaningful results for a population that really needs effective therapies.

And again, to try and address this issue that this was a small number of triple-negative patients on this trial. We did another exploratory analysis and presented this at ESMO Breast this year. We looked at the ER-low patients, so that’s the ER 1-10%. So ER low, HER2 low is very much aligned with triple-negative breast cancer. And so once again this doubles the experience, the number of patients with these TNBC/TNBC-like tumors, and once more you see striking activity in favor of T-DXd. So hopefully this gives clinicians a little more comfort and rationale for using T-DXd in the hormone-negative or hormone-low, HER2-low population.

We also looked at post-treatment therapies that the patients received, and we did calculate PFS2. And again, PFS2, the hazard ratio 0.51 for both hormone positive and all patients on trial. It’s very similar to PFS1, actually, so again, strikingly in favor of T-DXd.

You can see here the therapies in general that patients received post T-DXd. They were pretty well balanced and similar in the 2 arms of the trial. Very few patients got further ADC therapy, twice as many on the control arm as opposed to the T-DXd arm. But overall the post-trial therapies really don’t, I think, explain the significant survival advantage we saw with T-DXd.

So overall I think DESTINY-04 was an important study for the field. It was the first time we saw a HER2 targeted therapy work beyond HER2-positive breast cancer. This was the study that really established HER2 low as a targetable population and T-DXd as a standard of care for these patients. And this is important because that’s half of our patients with metastatic breast cancer.

So to end with an approach, an updated approach to sequencing, which is what we’ve been talking about already, for HER2-low metastatic breast cancer, first the hormone-positive patients. Again, I think most of us would agree we’d start with endocrine therapy first, and when these tumors become refractory to endocrine therapy we start to offer single-agent sequential chemotherapy traditionally, and after 1 line of treatment these are the candidates for trastuzumab deruxtecan today.

Following this we have the option to offer further single-agent chemotherapy to these patients, but today we also have the option to use sacituzumab govitecan, as we’ve been talking about, based on the TROPiCS-02 study. And it has a different antigen target, of course, but as you’ve heard the payload has a similar mechanism of action to T-DXd, so how effective this is going to be in sequencing is not totally clear. There are a lot of, I think, experiences people are reporting, retrospective experiences, I think Aditya, you’ve reported some. But we hope to learn more about how to properly sequence these drugs and whether it makes sense at all.

So for the hormone receptor-negative, HER2-low population, again, for those who have an actionable target that’s where we go first. So PD-L1 positive patients get immune checkpoint inhibitors. Germline BRCA patients we use PARP inhibitors. Eventually everyone gets back to chemotherapy.

In this case after 1 line of chemo I think we have 2 options, as we’ve heard about already, trastuzumab deruxtecan based on DESTINY-04 and sacituzumab based on the ASCENT trial. And truly they’re both options, but I think there’s a real sense that we have just more data for sacituzumab govitecan in this setting, and it’s pretty clear there’s probably a greater consensus to use that first. But there are options, and I think there may be some patients where we might consider T-DXd, particularly patients who have brain metastases might be better candidates for trastuzumab deruxtecan, where we know we see a lot of activity in the CNS space.

But these are individualize decisions. I think you look at the toxicity profile, patient characteristics, and preference. And it’s the same question what do you do after 1 ADC. Do you use the other one? This is a question for research today. And I’ll end there.

DR LOVE: Thanks, Shanu. Before we go on to talk about toxicities, Lisa, I’m kind of curious where you see dato-DXd fitting in. We talked about that Wednesday night here; we had a program on metastatic ER-positive disease, and I think Dr Sharma mentioned it as a consideration in the future. What are your thoughts about the data that was presented at ESMO on dato? And where do you see it heading? Any thoughts about indirectly comparing it to saci?

DR CAREY: I think it looks good. We don’t have survival data, so I think, again, we’re in the relatively data-free zone because it’s too early. But it looks good. It’s got a different payload, similar target, so you’re now getting a mixture of things. But it looks good. You’re seeing a nice improvement in PFS, maybe a little less ILD than you see with some of the others, with T-DXd.

DR LOVE: Peter, any thoughts about dato? And the other thing is they have a couple of toxicity issues that are kind of a little bit new for you. One, ophthalmic, which you guys haven’t gotten in, but everybody else is already in that one. And then stomatitis. Any thoughts, again, about that, Peter?

PROF SCHMID: So I think when we look at the differential use of dato-DXd, for example sacituzumab, we have to consider efficacy, we have to consider toxicity profile. The toxicity profiles of each ADC is unique, and you can’t say no TROP2 ADCs — there are no TROP2 ADC toxicities. It’s the same for HER2-targeted ADCs.

And so yes, there are subtle differences. GI toxicity, for example, more with sacituzumab. But based on the irinotecan derivative payload. On the other hand, more stomatitis or possibly dry eyes with dato-DXd.

When it comes to the efficacy question, at the moment all these ADCs have been used in a setting where patients had no prior ADC. And if you look at use of T-DXd and then possibly dato-DXd the change is in the target but not in the payload, and I think that needs to be taken into consideration. Taken on its own dato-DXd is a very impressive drug, but how do we fit this in in an environment where we want to give possibly T-DXd if a patient’s HER2 low, and would you after T-DXd prefer sacituzumab or dato-DXd? It’s a totally open question.

DR LOVE: I was mentioning last night we’ve been coming here for a long, long time. And actually I’ve never stayed in another hotel other than this one in like more than 40 years. This is like where I go every December, but really amazing to be here again and seeing all the data that’s coming in here today.

Identification and Management of Toxicities with Trastuzumab Deruxtecan — Prof Schmid

DR LOVE: I want to move on and talk about toxicity, and Peter’s going to go through this. This is one key question, and Aditya, one of the things that’s interesting is kind of the difference that you all got into the HER2 game, obviously, very, very early, then the GI people got in with the ToGA study with trastuzumab, but now T-DXd, again, you’re seeing in lung, upper GI, colon, et cetera. It’s very interesting how the different groups think through the risk of ILD and screening. You think you’re talking to 2 different people. Also HER2 testing, like the GI people constantly test, and then once it turns negative they stop using anti-HER therapy.

But anyhow, the general oncologists, we’re trying to process all this stuff. But of course the critical — one of the most critical issues is what does it take to stop T-DXd in terms of ILD, what grade. What we’ve been hearing from breast cancer investigators is Grade 1, they recover in 28 days, okay. Grade 2 or beyond do not. Not so strict in the other tumors. Any thoughts?

DR BARDIA: Yeah, absolutely. If you look at clinical guidelines and clinical trials, essentially patients who had Grade 2 ILD the recommendation was to stop T-DXd permanently. But in clinical practice when we use T-DXd, if you look at the definition of Grade 2, it’s symptomatic ILD. If a patient is having mild shortness of breath or cough or some other reason, is it truly Grade 2 ILD related to T-DXd? And if you don’t have any effective options, can you actually reintroduce T-DXd? So we have some anecdotal evidence that patients who have Grade 2 ILD on a case-by-case basis you could potentially consider T-DXd again rather than discontinuing it permanently.

So I think it depends on the specific patient, the other options they have, and how symptomatic they are related to the ILD.

DR LOVE: And always important, if possible, to bring the patient into these challenging decisions because certainly it’s their life.

Here’s some other questions we asked about tolerability issues with T-DXd. I think this kind of was a little bit of a surprise. I think people thought ADC, people weren’t going to have very much symptoms. So we see some alopecia. Again, you can see what the clinical experience is. Also in terms of fatigue, GI symptoms, pretty significant. You’ll hear comments from the virtual faculty as well.

We asked also, Shanu, and I’m curious what your thoughts are about this, in terms of chest imaging. For example, if a patient doesn’t need chest imaging are you going to do it anyhow looking for ILD? Most people say yes, and you can see — and again, when you go outside of breast then you start seeing this break down a little bit. But the breast people are sticking with regular imaging, even if they don’t need it, at a minimum of 12 weeks. Is that the way you do it, Shanu?

DR MODI: Yeah. I mean in the clinical trials we were doing the chest imaging every 6 weeks. Remember, that was to look for efficacy. We weren’t actually doing that for lung toxicity, but we were able to look at it.

DR LOVE: I didn’t know that. That’s interesting. Really? That’s interesting. So you were looking for efficacy. You weren’t even looking for ILD.

DR MODI: Yeah. This was not for ILD. We were doing that for efficacy.

DR LOVE: So you didn’t know when you started that it caused ILD?

DR MODI: That wasn’t the primary reason for doing the CT scans every 6 weeks.

DR LOVE: Wow. Interesting.

DR MODI: Thank goodness we did.

DR LOVE: Yeah.

DR MODI: But it’s hard to get payors to cover a CT scan every 6 weeks. Most of us can get it covered every 9 to 12 weeks, and so I usually recommend that. People do scan their patients. I think what worries me is people who don’t scan their patients for 6 months. That’s a problem. So I’ve found creative ways, and it depends, a lot of it, on my threshold of what that patient’s risk is also for lung toxicity. If they have other risk factors sometimes I’ll do a scan in between their staging PET scan, so they’re alternating, but we’re doing some imaging roughly every 8 to 9 weeks of the lungs in these patients.

DR LOVE: Any other non-imaging, like pulmonary functions or O2 sat monitors? It doesn’t look like many people do that.

DR MODI: Not routinely. We didn’t do that in the studies as a way of screening patients. I mean there are tools that are being developed to monitor oxygen saturation at home. So hopefully in the future we are going to have some non-cumbersome ways to monitor patients besides just chest imaging.

DR LOVE: What about this one, Shanu, do you see more ILD in Asian people even if they’ve immigrated here? It looks like there’s a consensus in this faculty. Just kidding.

DR MODI: I’m trying to remember. I mean I don’t believe we saw that that was a consideration in the American patients, that it was the Asian American patients who had more ILD, but in patients who were enrolled, in the early trials anyway, remember a lot of them were enrolled first in Japan, so we saw more ILD in those patients. And I think we know historically Japanese patients do tend to get more lung toxicity from some of the therapies we use in general, so I it is considered a potential risk factor to be receiving this therapy in a Japanese population.

DR LOVE: Interesting. So Peter, I’d like you to comment on Dr Tarantino’s thoughts about the day-to-day management and particularly some of the acute tolerability issues. You’re going to go through toxicity. Of course people always think about ILD. Here’s Dr Tarantino.

DR TARANTINO: I would say we were not really expecting these chemo-related side effects at the beginning when we started using it, but we rapidly understood that T-DXd is more chemo than T-DM1, both as the molecule and also the feeling of the patient, and especially GI side effects. In all the trials we see up to 70% rate of nausea, vomiting, and we saw that in clinical practice. But we rapidly escalated to at least 2-drug regimens up front, serotonin antagonist and dexamethasone, and also adding an NK-1 antagonist for most of the patients.

I utilize at 3-drug regimen for prophylaxis for most patients and usually I also utilize ondansetron for 3 to 4 days after the infusion. And for some patients we have to advise that it can occur also later in the treatment course. We know that these ADCs have longer kinetics than traditional chemo, and so it’s quite common to see delayed nausea, and we do know that we have to advise/educate patients about this.

The NCCN Guidelines recommend that because they suggest that T-DXd is highly emetogenic, and I think it’s a good idea for most patients.

DR LOVE: Any comments, Peter? And any thoughts about this thing that maybe with ADCs it occurs later?

PROF SCHMID: Yeah. So I think, again, we had a steep learning curve with those new ADCs, and I often say to patients if you put on the naked antibody here and single-agent chemotherapy there, the first/second generation of ADCs, we’re a bit closer to 2 antibodies. Now we are much closer to single-agent chemotherapy, and that’s probably ultimately observed. But when these drugs were developed in the clinic we didn’t do our normal homework from day 1.

For example, as Paolo mentioned, we didn’t have effective antiemetic therapy, and that’s if you treat patients with antiemetic prophylaxis, nausea is very well manageable. But if you don’t, of course patients have much more nausea then we would have expected with the older generation of ADCs. So it comes down to prophylaxis and intervention. If you do this well then it’s well manageable.

The delayed nausea comes from a slightly different probably pharmacokinetic behavior and a different release over time. So these drugs are not just in and out of the body. They can stay in your body for a slightly longer time.

DR LOVE: So Lisa, I’d like to see whether you identify with what Dr Brufsky has to say.

DR BRUFSKY: First of all, I think the amount of nausea that we got in clinical, in real life now really kind of matches and if not exceeds what we saw in the trial. I think that really pushing the anti-nausea medicine up-front. And we’re going to see some really interesting data I think on olanzapine actually. At San Antonio I just saw the titles of the abstracts.

In terms of the ILD, I generally will do a CT. I don’t do them every 6 weeks like my colleagues. I’ll still do them every 3 months. But the real question that I think a lot of us who use a lot of T-DXd now are running into is you have someone who’s responding dramatically to the drug and you see some haziness on their CT and you have no idea what it is and yet you really, really, really want to continue them and it’s really, really hard to say, this is ILD and I want to, I have to stop the drug and potentially put them on steroids for, a steroid taper for a month and rescan them. You have to do that. That’s the hardest thing as an oncologist to do I think.

This is really the central question is you see this thing, you really don’t know what it is, the patient’s responding dramatically, what do you do? That’s a toughie and we’ve got to stop. You just got to say you have to stop. And you have to think about steroids. Some give it for asymptomatic, some don’t, but that to me is the biggest one of all.

DR CAREY: Yeah. I mean I think he’s outlined it very well. The truth is that I think routine scanning is important because if you have asymptomatic ILD it resolves you can retry it. If you get to symptomatic then typically we do not.

And I think Adam also outlined maybe the 1 place where if you ask a very specific question, as you always do, and you say okay, they’ve got Grade 2, they’ve got symptomatic ILD, are you going to rechallenge? Would you be willing to? And everybody says no. But if you said okay, they had some hazy stuff, and they also had a fever, and then it got better, would you retry it? And when there’s clinical uncertainty about whether something was actually a drug-induced change in the lungs or whether it was something else then I think many of us would try it again.

DR LOVE: So I’m going to hold off on the next video to make sure we have enough time, but I want to make sure Peter has enough time to go through the data now that we’ve kind of talked a little bit about your own clinical approaches to these patients. Now let’s come back to what we know from a very, very objective data point of view. Peter?

PROF SCHMID: Yeah. Thank you, Neil. So if you look at what we know about trastuzumab in HER2-low disease (T-DXd) you can see it is essentially single-agent chemotherapy. It has a distinct profile, and that’s slightly different to the chemotherapies that were used in the treatment of physicians’ control arm. You see here in this trial nausea is the most common side effect, fatigue, and then going down. Myelosuppression, for example, slightly less than in the control arm. Diarrhea slightly more than in the control arm.

The AEs of special interest are ILD, and we’ll talk about this in a minute, but also we haven’t talked about left ventricular dysfunction, and that obviously comes from HER2 as the target. But if you look at the percentage of patients that have LVF changes that’s relatively low, and in my personal experience not a significant clinical concern. The most common AE associated with treatment discontinuation was clearly ILD that we already spoke about, but for that reason to avoid discontinuation I think it’s really important to recognize this early and intervene relatively early.

Now this is an attempt to just put next to each other, we always have to be careful when we compare drug A with drug B, but 2 ADCs, both targeting HER2, but obviously different payload, different linker, different generation. And again you can see these 2 ADCs, T-DM1 on the left, T-DXd on the right, have just a very distinct side effect profile. You see on the right side with T-DXd more ILD than with T-DM1, where it was also observed. You see on the other hand, on the left side with T-DM1, for example more peripheral neuropathy, that which really isn’t a feature with T-DXd. And then nausea as one of the manageable side effects, again, in the early trials of T-DXd to a higher degree than with T-DM1 but can be well managed.

And if you summarize what drives ADC toxicity then it is a mix. It’s a mix of target and the tissue distribution. It’s the payload. It’s linker. It can also be direct antibody effect on target or immune effect, and then possibly patient factors, for example when it comes to ILD it’s important to know what is the lung history for a specific patient.

If you go through some of the side effects, we already spoke briefly about management of nausea, so I don’t want to go into this in too much detail, but I think the key message here is treat it as a moderately emetogenic drug and therefore we would use at least doublet or possibly ideally triplet emetogenic prophylaxis with 5HT3, steroids, and possibly NK-1 antagonist. Again, the delayed nausea is probably a little bit more pronounced than with some of the normal chemotherapy, not ADC-based chemotherapy, and what works in our experience fairly well is actually for delayed nausea 5HT3, which we would normally not use with standard chemotherapy. There’s a number of drugs we can use for breakthrough nausea.

If you look at myelosuppression, I don’t think that’s very different to standard chemotherapy drugs. The normal recommendations are wait for ANC to be above 1 and platelets recovered. We wouldn’t give primary G-CSF prophylaxis unless in risk groups, but of course secondary G-CSF prophylaxis as indicated. The dose reductions you can see in the top right corner. First reduction from 5.4 down to 4.4, and then 3.2, and if you still can’t manage the side effects then patients may have to discontinue with their therapy.

LVEF changes, similar, but a little bit different to what we know from, for example, trastuzumab, and in clinical practice less of a problem. You can see when you look at, as we have known for a long time for HER2-targeted therapies, you look at the absolute LVEF, but we also look at the change from baseline, and depending on how low it has dropped and how much it has dropped the drug is either held or possibly in some situations discontinued.

I wanted to talk a bit more about ILD. You’ve already heard the definition is it’s basically a mixed bag for diffuse parenchymal lung disorders, and that includes but goes beyond pneumonitis. If you look at the symptoms, very nonspecific, it can be cough, it can be fever, it can be shortness of breath, and that’s tricky because patients with advanced cancers can be slightly short of breath, again they have a cough. And that’s where we need to work out is this ILD triggered or is it possibly cancer triggered. That plays a role when you decide whether it’s Grade 1 or Grade 2 in terms of the tox grading.

The clinical signs, there can be inflammation. Obviously it’s largely dependent on the chest x-ray or better on the CT scan. But we can also sometimes see changes in pulmonary function tests. Risk factors in particular if patients have a history of lung changes, ILD, pneumonitis, smokers, possibly age, but I think the main point to look out for is history of lung problems.

If you look at when does ILD occur and it seems that the median time to onset, really mixed bag studies across all tumor types, was 5.4 months. And if you look at the incidence it was here around 15% in some of the studies. It’s probably between 10 and 15%. But if you look at the grading, the majority where Grade 1 or Grade 2. I think the key job for us as clinicians is to catch it when it’s Grade 1 and stop it from getting to Grade 2 because according to the current guidelines if it’s Grade 2 then obviously you have to discontinue, and of course under all circumstances we want to avoid Grade 3 or 4 or higher.

Again, the latest data from the DESTINY-Breast04 study show the data specifically for HER2 low, and you can see the median time to onset 129 days. It’s a little bit later than what was seen in the control arm, where obviously in a very small number of patients ILD was observed.

So what do we do? We should monitor. We should monitor for symptoms. We should review the patient’s regularly. It’s not a drug where you say come back in 6 months’ time. We can, I don’t think that’s routine practice in most centers, but you can if you want to monitor the oxygen saturation, and as you have already heard, do a CT scan probably every 9 to 12 weeks. If an ILD is suspected lung function testing, CT chest scan. Bronchoscopy is really there to exclude other reasons if you’re not quite sure, and the normal workup that includes infectious causes.

Now in terms of how do we manage it the key thing is to work out is it Grade 1 or higher, and the definition is based on whether it’s symptomatic or asymptomatic. If it’s Grade 1 hold the drug. You can restart the drug if it has resolved within 49 days, at the same dose if it has resolved within 28 days, lower the dose if it took a little bit longer. We would often give steroids relatively early on, and I think that’s probably the right thing to do. With Grade 1 it can be orally given at 0.5 mg.

When it comes to Grade 2 we should permanently discontinue the drug. We should give steroids. You should taper over a long period of time, and those patients who don’t improve we need to think about escalation of the immune suppression.

DR LOVE: Awesome. So we’re getting some great questions from the audience as well. And Shanu, I just had a flashback to an interview I did with you right after you presented the data, and you said something to me, and I’m going to bring it up in a second. But I have a question here from Marco who says, “Given the success of combination chemotherapy what about ADCs with more than 1 chemo payload?” I don’t know if that’s ever even been attempted. But what I remember you telling me about was “ADCs where the payload’s not chemo, but some kind of immune thing.”

DR MODI: Right.

DR LOVE: And then I heard somebody present, I don’t think it was in breast, somebody presented something. I said, “Oh, that sounds like what you talked about.” So like what’s going on in terms of new development of ADCs? Can you put 2 things on there, 2 different chemos?

DR MODI: Yeah. I mean it’s the Wild West coming back with ADCs. So these are second-generation drugs we’re talking about, but I mean the future generation drugs in the pipeline are really exciting. I mean they are looking at different payloads, different types of payloads, so immune-stimulating payloads. There are ADCs where it’s not chemo, but we’re linking something that creates a T-cell response against the cancer. Speaking specifically about the BDC-1001 probably that’s what you saw, and that drug is moving into a Phase II trial in breast cancer.

So yeah. I mean we’re seeing radioisotope payloads and there is potential, I think, to combine 2 different ADCs with different payloads, but also to create an ADC that has different payloads, different types of payloads attached to it. So all of this is possible, and there are people interested in developing or advancing the technology and the field in this direction. So I say stay tuned. I think there’s a lot more we’re going to be seeing with ADC therapy in the future.

DR LOVE: So Lisa, Bertha wants to know how does the presence of brain mets alter your algorithms, the algorithms we’ve been talking about tonight? That was her general question. Any scenarios? What she’s talking about, I think, are situations where, triple negative for example, where you all are saying saci first. Does it change if there are brain mets?

DR CAREY: Yeah. It’s a great question. There are fewer data than there are in the HER2-positive space in this regard, but triple negative also gets brain mets quite frequently. And one of the differences in triple negative is that typically the brain mets are associated with also progressive disease in the systemic thing, so you really have to — we’re much more dependent on systemic approaches than locoregionally directed ones.

Fortunately right now there have been increasing amounts of data, some were presented at ESMO, of T-DXd and its efficacy with active brain mets. There’s not a ton of data, but it’s very encouraging that it in fact does have reasonable effectiveness in that setting. So personally I think there’s a little more data for it, so if I’m in that, what Shanu had, the jostling between the ADCs, that might tip me over to reaching for T-DXd.

Future Directions in the Management of HER2-Low Breast Cancer — Dr Bardia

DR LOVE: So in a minute Aditya’s going to go through some of the new directions that we’re looking into, but one of the other things we asked is we asked all of these investigators to talk about the last patient they treated with T-DXd with HER2-low metastatic disease. So you can see this is a summary. In real life, almost all of these patients were hormone receptor positive, which is the case here. And you can see it’s split between IHC 1 and 2+, although I don’t know that that makes a big difference.

And then in terms of response most of these patients, more than half, it looks like more than half responded. Most tolerated well. I think this is Virginia’s case that she talked about Wednesday night of a Grade 3 ILD, so one of those patients did get ILD.

And Peter, I’ll be curious what your thoughts are about Dr Pegram’s speculations for the future.

DR PEGRAM: I’m much more excited, Neil, as you might tell, from the use of ADC technology beyond the cytotoxic payload fire drill.

We need more cytotoxic payloads.

What we need in our toolkit as oncologists, we need taxane payloads, we need alkylator payloads, maybe even anthracycline payloads. We need the active drug classes for which we have curative therapies already to be available hopefully with more efficacy and less toxicity than using chemo in the next generation.

So we need to be more creative, in my opinion, and this is just philosophy here.

In terms of resource allocation, I would much rather see a different payload with that target, or a different target with that payload would be of more interest to me personally.

DR LOVE: Any thoughts?

PROF SCHMID: Yeah. I think Mark is absolutely right, and this is what was said earlier. It doesn’t make sense to always use a taxane therapy. An ADC is a targeted delivery device, but we still need to deliver the right target, and I think the next generation will focus a little bit more on this, to have more diversity in terms of the payloads we are using, whether it’s chemotherapy based but also are there other drugs from an immune to, for example, radioisotopes as we heard before, or combinations of different payloads. I think that’s where the developed is going.

DR LOVE: So one more question that I asked Mark, and Lisa I’ll be curious, before Aditya goes through some of these data on HER2-low, sneak into HER2-positive metastatic disease and ask you the question that you’re about to hear Mark answer, but I’m going to ask you first, Lisa. How do you think we’re going to be treating first-line metastatic HER2-positive disease in 3 or 4 years?

DR CAREY: First-line HER2 —

DR LOVE: HER2-positive mets.

DR CAREY: Oh, I think we’re going to be moving the ADCs. They’re going to keep marching up. I mean I think the — listen, we’ve had the CLEOPATRA regimen as the first line for a number of years because it had such a stunning survival advantage so quickly, but it is an old-fashioned way of giving therapy. And so my anticipation is that we will have — T-DXd would be my guess is going to be the first line.

DR LOVE: Do you think that pertuzumab’s going to add to T-DXd?

DR CAREY: I don’t know. I think it might. I’m struggling a little bit to think exactly what’s going to happen at the level of the receptor when you have the trastuzumab deruxtecan and you’re also adding a free anti-HER2 antibody, whether that’s going to be a good thing or if that’s going to start becoming a bit of a log jam. I think the small molecules here might also be possibilities, so it’ll be interesting to see, although the toxicities may be difficult to manage together.

DR LOVE: Here are Dr Pegram’s thoughts on that.

How do you think we’re going to be treating first-line therapy of metastatic HER2-positive disease in 3 or 4 years?

DR PEGRAM: Well, there are a of threats to CLEOPATRA that are ongoing as we speak, not the least of which is trastuzumab deruxtecan. The DESTINY-Breast09 trial will put T-DXd head-to-head against the CLEOPATRA regimen of taxane/pertuzumab/trastuzumab. And there’s a third arm in that trial also, which is very interesting, which is T-DXd plus pertuzumab, so a 3-arm randomized trial. I suspect that the T-DXd arms will probably both beat trastuzumab/pertuzumab and the taxane. That’s my gestalt, but I’m very anxious to see that study.

Meanwhile, there’s an ongoing trial in those HER2-positive patients in the first line that are also ER positive. They’re randomized to maintenance CDK4/6 inhibition with palbociclib versus not. That’s the PATINA trial. That should read out very soon. There’s also an alpelisib trial in the maintenance phase given with the 2 antibodies after the taxane sequence in a CLEOPATRA-like regimen, and there’s also a tucatinib maintenance trial that’s ongoing with the 2 antibodies in a maintenance fashion after you finish the taxane and CLEOPATRA. So any or all of those might beat CLEOPATRA at the end of the day, so there probably will be big changes in the first line well before 5 years from now, probably in the next year.

DR LOVE: So Aditya, you put together a talk on this topic. Before you start, anything you want to say about what Mark just said?

DR BARDIA: I fully agree. I think all of these are potential therapies that could change practice, all of the maintenance approach would be extending CLEOPATRA. So if the alpelisib trial or the CDK4/6 trial, or the tucatinib trial is positive it’s like an addendum to CLEOPATRA, but the base remains CLEOPATRA.

DR LOVE: I’m just kind of curious, Peter. Do you use CDK in your HER2-positive patients? Do you use CDK inhibitors in your HER2-positive patients? CDK.

PROF SCHMID: CDK. Sorry. I’m half deaf.

DR LOVE: Sorry.

PROF SCHMID: I think the role of endocrine therapy is a little bit difficult in HER2-positive breast cancer. I think it will work, but currently fitting it into our treatment standards is difficult. We are just changing our first-line therapy probably over to an ADC-based therapy with T-DXd.

I think the concept of adding it in as a maintenance therapy is interesting, giving endocrine therapy with HER2-targeted therapy on its own has never really fully taken off. So I think it’s probably a path where we are not utilizing as much as we should do.

DR LOVE: All right. Aditya?

DR BARDIA: Absolutely. So I’ll review future directions in HER2-low breast cancer with the focus mostly on ultralow, so we’ll start with HER2 ultralow, the rationale for T-DXd in that setting, review clinical data, including the DAISY trial, review upcoming clinical trials in this setting, both metastatic and localized, briefly talk about combinations of T-DXd with other therapies, particular immunotherapy, and then finally other HER2 agents besides T-DXd.

So if we look at the definition of HER2-low and ultralow breast cancer we already reviewed HER2-low breast cancer. Ultralow is a subset of IHC 0. If you look at the definition of HER2 IHC 0 it’s no staining or membrane staining that is incomplete in less than 10% of tumor cells. So even HER2 IHC 0 is not 0. Some cells can have HER2 expression, and that’s the definition of HER2 ultralow. The reason this is important is because of antibody-drug conjugates which can deliver the payload if there is some HER2 expression, and then because of the bystander effect can impact other cells that do not express the antigen.

We saw the results from the DAISY trial, which looked at activity of T-DXd based on the expression of HER2. We saw very high response rates in HER2 IHC 3, or HER2-positive tumors, lower response rate in HER2 low, and then a bit lower in HER2 IHC 0, but there was still some patients who had responses, including complete responses, even with HER2 IHC 0.

So this has led to interest in evaluating T-DXd in this setting. And an example of this is the DESTINY-Breast06 trial. This is looking at T-DXd versus investigator choice of chemotherapy in patients with endocrine-resistant ER-positive breast cancer who’ve not received prior chemotherapy, so it’s like a first-line chemotherapy trial. This trial also includes patients with HER2 IHC 0+, or HER2 ultralow. So it’ll be interesting to see activity of T-DXd in this setting. The trial has completed enrollment, so hopefully we’ll see the results next year.

Then finally how about early breast cancer with T-DXd? The TALENT trial looked at T-DXd with or without endocrine therapy for patients with hormone receptor-positive, HER2-low breast cancer, Stage II/Stage III. Patients in the trial received T-DXd for 6 cycles, it was amended later on to 8 cycles, with or without endocrine therapy.

In terms of response rate you saw that the use of T-DXd was associated with radiological response based on ultrasound assessment before treatment and then after the completion of T-DXd therapy, but the rate of pCR was not very high. So it looks like there is decrease in the tumor burden, but it does not translate to pathological complete response.

And then in terms of change in HER2 it was interesting to see that with the use of the T-DXd as antibody-drug conjugate there was change in HER2 expression, as well, with some tumors that were HER2 IHC 1+, 2+ changing to HER2 IHC 0. For some patients the HER2 expression did not change, and then rarely it went the other way, where the tumors that were HER2 IHC 1+ it increased to HER2 IHC 2+ at the time of surgery.

How about immunotherapy? There’s interest in combining immunotherapy with antibody-drug conjugate. As was mentioned by Peter previously essentially antibody-drug conjugates are chemotherapy. They’re delivering chemotherapy selectively to tumor cells. So when the antibody-drug conjugate delivers the payload it causes apoptosis of the cancer cell, which can then release antigens that can activate the immune system in the microenvironment, which can then result in expression of PD-L1. So a combination of immunotherapy with antibody-drug conjugate potentially could be synergistic.

We’ve seen some data combining T-DXd with immunotherapy, such as the U105 trial where patients with HER2-positive breast cancer received T-DXd plus nivolumab, and you saw a response rate that was similar to what has been seen with T-DXd alone. The trial also had a HER2-low population, and again the response rate was similar to what was seen with T-DXd alone. This was a single-arm study, but if you compare these results to historical data from T-DXd the response rate was similar.

So it’s unclear at this time whether immunotherapy adds to the efficacy of T-DXd, but we don’t have randomized data. So ultimately we need randomized results before we can conclude whether immunotherapy adds to the efficacy of antibody-drug conjugates like T-DXd.

How about other ADCs? There are other ADCs targeting HER2. An example is trastuzumab duocarmazine. It targets HER2 but has a different payload. It’s an alkylating payload. Duocarmazine has shown activity in HER2-positive as well as HER2-low breast cancer.

The problem with the drug is the safety profile. It tends to cause a high degree of keratitis and eye toxicity. And while the Phase III trial investigating trastuzumab duocarmazine versus standard therapy in HER2-positive breast cancer was positive, because of the safety profile it received a CRL.

Another drug is disitamab vedotin. Here a different antibody targeting HER2. It’s not trastuzumab, it’s disitamab, and with a different payload which is vedotin, has shown good preclinical results, ongoing investigations evaluating this agent, both in HER2-low, as well as HER2-positive breast cancer.

And then finally sacituzumab govitecan, which does not target HER2. It targets TROP2. But given the classification of HER2 low there’s been interest in looking at efficacy of sacituzumab govitecan by HER2 status, the hypothesis being that HER2 status should not impact the efficacy of sacituzumab govitecan given that it targets TROP2, and that is what was seen in the TROPiCS-02 study, as well, that both in HER2 low, as well as HER2 IHC 0, the benefit with sacituzumab govitecan was maintained over chemotherapy.

And besides this there are other antibody-drug conjugates in development. The question we’re already facing now that will only increase in the future would be how do we sequence the different antibody-drug conjugates. We’ll have to pay attention to both the target and the payload. And I think ultimately we need to figure out what’s driving the resistance to ADC-1 so that we can choose ADC-2 based on mechanisms to avoid cross resistance. So if the resistance to ADC-1 is predominantly because of the payload using an ADC-2 with a similar payload would likely result in cross resistance.

And example of this was seen with sacituzumab govitecan where mutations in both TOP1 and TROP2 were described in a single patient, and this has implications for sequencing because if the resistance is predominantly to the antibody then using say trastuzumab deruxtecan followed by trastuzumab say duocarmazine or vedotin would likely result in resistance, but if you use trastuzumab deruxtecan and then after that use datopotamab deruxtecan or even sacituzumab govitecan likely there will be efficacy because the resistance is to the antibody. And the opposite if there’s resistance to the payload.

Now this is nice in principle. Ultimately how we’ll incorporate this into clinical practice, we’ll need biomarkers, ways to measure whether the resistance is predominantly because of the antibody payload. So a lot of research needs to be done before this can be incorporated in clinical practice.

So in summary, even HER2 IHC 0 has some HER2 expression which can be leveraged as a target for ADCs with bystander effect. Trastuzumab deruxtecan is approved for HER2-low breast cancer and is being evaluated for ultralow as well. ADCs with immunotherapy have demonstrated efficacy, but how much the immunotherapy adds to the efficacy to ADC is currently unknown but being evaluated in clinical trials. And given the multiple HER2 ADCs in development we need better understanding of resistance, whether it’s antibody mediated or payload mediated, to guide the sequencing of ADCs.

DR LOVE: So we’re getting a bunch of questions that we’ll finish out with. So Shanu, Nicholas wants to know how often you do echocardiograms with T-DXd, and what is the role, if any, of ADCs in leptomeningeal disease?

DR MODI: Good questions. So as you saw in HER2-positive breast cancer of course we’re used to following ejection fraction. So this is a HER2-targeted therapy, so again we do follow the echocardiograms or MUGAs every 3 months on these patients who are getting T-DXd in the HER2-low setting.

In terms of leptomeningeal disease, I think it was today or tomorrow that the DEBRA trial is going to report the activity of T-DXd specifically for patients with leptomeningeal carcinomatosis. We’ve already seen small case reports where there’s clear activity. I want to say it was an MGH or Dana-Farber study where they — or a group study.

DR CAREY: TBCRC.

DR MODI: TBCRC. They showed I think it was 8 patients. Half of them had a response to T-DXd, and these are patients with leptomeningeal disease. So something to be really excited about, systemic therapy for this really challenging population. So I’m actually as excited about the CNS benefits of T-DXd as I am about its systemic benefits.

DR LOVE: Interesting. So Aditya, Rajalaxmi, Dr McKenna from our crack chat room — we have the best chat room. They ask great questions. So I love this one. She has a patient with Stage I ILD but she’s been measuring the O2 sat. So the patient’s asymptomatic, Stage I, but has a drop in O2 sat.

DR BARDIA: So I would hold T-DXd because even for Grade 1 pneumonitis the recommendation is to hold T-DXd and give steroids. But I would expect that in this patient the pneumonitis would improve, and then you can use T-DXd again.

DR LOVE: But I guess the issue here is like as you’re treating this patient would you retreat the patient? In other words, would you consider the drop in O2 symptomatic even though the patient didn’t experience it?

DR BARDIA: I would not consider the drop in O2 symptomatic per se. It also depends on the amount of disease the patient has and how urgent the need for T-DXd is, but in general I would feel comfortable treating this patient if otherwise the patient is feeling well.

DR LOVE: So Lisa, a lot of questions about IO plus ADCs, and of course we’re seeing that in other cancers. Particularly in bladder cancer they have an IO plus ADC, so pembro and enfortumab first line, and it was better than platinum-containing chemo. First it was approved in older people, and I’m like the younger people want this thing too. Any thoughts? I mean obviously IO plus chemo in general, in lung cancer, is a very common strategy. You would think IO plus ADC makes sense. Any thoughts?

DR CAREY: Well it does, and there’s some preclinical data and some modeling that suggests that in fact IO would be synergistic with ADCs in a particularly nice way. I mean you do do better giving a cytotoxic with IO, particularly in breast cancer, and we actually have a trial in triple-negative residual disease, who have received KEYNOTE-522 and still have residual disease, they continue their pembro, and they’re randomized to receive sacituzumab or not. That’s an Alliance trial.

DR LOVE: Peter?

PROF SCHMID: Yeah. We presented data actually a few weeks ago, the BEGONIA trial, where we combined ADC with IO, and one of the arms with dato-DXd I think was particularly interesting. There was a group of patients with first-line TNBC, predominantly PD-L1 negative, so nearly 90% of patients were PD-L1 negative. If you look at what we saw in Phase III trials with chemo and immune therapy there’s a response of 40-45%, PFS of about 5 months. Now with dato-DXd, an immune therapy, in that cohort we had a response rate of nearly 80%. We had a progression-free survival of 30.8 months. We had a duration of response of 15 months. Now that’s a very powerful signal, which I personally don’t think it’s just the ADC. I think there’s an interaction with the immune checkpoint inhibitor, and it raises the question whether it can widen the target indication and actually make immune therapy work in PD-L1-negative patients.

DR LOVE: That was durvalumab in that study?

PROF SCHMID: That was with durvalumab in that trial, yes.

DR LOVE: Interesting. So I want to thank the faculty. Thank you for attending. We’re heading out to ASH. Check us out tomorrow. Have a great evening. Thanks so much.