This transcript has been edited for clarity. For more episodes, download the Medscape app or subscribe to the podcast on Apple Podcasts, Spotify, or your preferred podcast provider.
Kathy D. Miller, MD: Hi. I'm Dr Kathy Miller. Welcome to the Medscape InDiscussion podcast series on breast cancer and HER2. Today we're going to talk about HER2 testing positive, low, ultralow, and all of the gradations in between. I am delighted to introduce my guest today, Dr Sunil Badve. Dr Badve is the vice chair of the Pathology Cancer Program and a professor in the Department of Pathology and Laboratory Medicine at Emory University School of Medicine in Atlanta.
Dr Badve is a well-known breast surgical pathologist and translational researcher with expertise in breast cancer and thymic pathology, and I know he's thought deeply about this topic. Sunil, thank you for giving us your time today.
Sunil S. Badve, MD: My pleasure, Kathy. It's wonderful to join you for any podcast or meeting.
Miller: I'm going to have you join me in the wayback machine to the early days of HER2-targeted therapy and HER2 testing. This predated the American Society of Clinical Oncology (ASCO)-College of American Pathologists (CAP) initiative, and I remember spirited discussions among pathologists as to whether immunohistochemistry (IHC) was more reliable, whether fluorescence in situ hybridization (FISH) was more reliable, which was a better predictor? What should we consider positive? As somebody trying to make sense of pathology reports and make treatment decisions, it really felt like chaos and uncertainty.
Badve: It's been chaotic for a long period of time, and some people will argue that it is chaotic, even today. So let's start at the beginning, right? In the late eighties, with the discovery of genes that could cause cancer, there was a search to find those oncogenes in cancer. And, as is well-known, Dennis Slamon's group, along with Mike Press, identified HER2 as an analogue that was associated with breast cancer. The HER2 name is later given, but the gene name is ERBB2 (Erb-B2 receptor tyrosine kinase 2) gene — again, a homage to the oncogene theory.
A couple of things happened around the same time. Mark Greene's group from UPENN identified an antibody, working with the neuroblastoma models, and therefore the neu, which was basically an HER2-inactivating antibody, and they could stop the growth of tumors.
Now we had a perfect storm of a gene identified in breast cancer, a preexisting antibody that could block the impact of the particular gene mutation or expression. And that led to the development of Herceptin. And then the question came up: What is the best way to test these particular proteins?
Early work from Press' lab showed that Western blot analysis and similar molecular methods were far better at detecting gene expression than IHC. Of note, even today, the frozen-section analysis of normal breast tissue showed some levels of HER2 protein expression.
So then we come back to the clinical trials assay, where there were two different antibodies that were developed. From there we went to the HercepTest, which was a polyclonal antibody. Now, the concordance between the clinical trials assay that was used in the metastatic set data vs the Hercep antibody was best at the 10% cutoff, and that's where I'm told the 10% cutoff comes in.
There's no other evidence to say that 10% is a magic number of any sort. There's nothing different between 9 and 11, but the concordance was best between the two assays at the 10%, and that's the reason we started using the 10% cutoff and the intensity of staining. And the completeness of staining based on that.
So the one, two, three, all of that was based on the 10% cutoff. And then, whether the membranes were complete or not and the intensity of the staining. That's really the history of the assay.
Miller: That was part of the history that I didn't know where the 10% came from. I do remember two different antibodies. And I remember at the time, there were some thinking FISH testing, looking for too many copies of the gene. Very simplistically, assuming that if you have too many copies of the gene, you have too much of the protein. And to a non-pathologist, it felt like FISH testing should be more reproducible because you're counting dots rather than qualitatively trying to estimate is this brown stain dark chocolate brown or is it milk chocolate? Is it barely beige? And what proportion of the cell surface is staining?
It felt to the non-pathologists in the room like FISH testing is going to be the better way to go. I know there were some laboratories in those days that abandoned IHC and just sent everything for FISH testing. In this chaos, the ASCO and CAP came together and took a close look at this. I remember being surprised, and not in a good way, that FISH testing was not more reproducible.
Badve: It's a classic example of looking at upper management to see what's happening on the ground. The DNA molecules that we are looking at in the nucleus are upper management in this case, which leads to translation to RNA, which leads to the development of proteins, and then the proteins have to be there on the cell surface and retain for a long enough time for the IHC to work.
That's really the major reason why we have discordances between the labor force, which is the protein, and the upper management, which is the DNA. Obviously, we want the protein on the cell surface for the antibody to bind, and therefore, you want the IHC to be positive. On the other hand, the thought process was the upper management, the DNA alterations indicate a real change within the cancer cell, and maybe that is far more important than what is happening at the cell level because the protein may be there in lesser or greater amounts, it doesn't really matter as much. That debate continued for a long period of time, and it got supplanted by the whole story of HER2 because now we are increasingly recognizing that the protein on the cell surface is far more important than what is happening genomically.
Miller: People always revert to proteins as where the work is done, regardless of what's happening in the genes. And at least for the clinicians, deciphering the pathology reports got easier after the ASCO-CAP guidelines. I think we saw less variability. The results felt more like things you could hang your hat on for making treatment decisions when there were some quality control guidelines of how tissue should be handled. How long in the fixative, specific criteria, for what is still subjective to at least some extent from the pathologist's eyes. The one thing that ASCO-CAP did was not invent but use my now least favorite word, equivocal, in their designation because that just created so much confusion. And around the same time, patients more frequently got copies of their pathology reports, and many of our patients interpreted the equivocal designation as a throwing up of hands, they don't know anything. They just don't know. And I know that's not what they meant to indicate.
Badve: When the data are clear-cut, we don't need consensus statements. When the data are muddy or nonexistent, we need to have a consensus as to what might be the best. So the consensus statement reflects not only our state of knowledge but also our state of ignorance. And equivocal was a category that was invented to essentially hide the state of ignorance. But at the same time, Herceptin was a new drug. There was a lot of worry about cardiac toxicity. A lot of those things were coming to a head, so people were like, well, if you are not sure, let's step back a second and redo the test with another method just to confirm everything is nice, everything is wonderful, and yes, this patient really needs to be treated with that particular form of therapy. So that was the basis of using the category of equivocal, saying we are not quite sure, it's very close to the borderline. Let's retest. And if you think of equivocal in that particular way, it makes a lot of sense.
Miller: That's how I always interpreted it — that the underlying biology is a continuum, and we've put these arbitrary fences. It seems reasonable to say that if you get a first result that is near an arbitrary fence, you should take a closer look, whether that's repeat the testing, use a different technology, or send it to a different laboratory. You should take some time to make certain you clearly know what side of the fence you're on before making a treatment decision. So I think conceptually that makes sense. It was just a poor choice of words that led to a lot of confusion.
Badve: I absolutely agree. In hindsight, a lot of these things become clear because it's now 10-15 years since the first ASCO-CAP guidelines. So the first ASCO-CAP guideline was, be careful, guys. I think that's a simple message that one can understand. By the time the second ASCO-CAP guideline version came along, it was, yeah, the drug is safe. We don't need to be that worried. Yeah, let's go ahead and give it to as many patients as we think might be eligible. We don't need to be that cautious. And that's the reason the 30% IHC cutoff that was put in the first guideline came back down to the FDA-approved cutoff of 10%.
Miller: So things have gotten more confusing with the antibody-drug conjugates (ADCs) because, for many of us, the goal was identifying those that were truly HER2 driven, who were likely to benefit from trastuzumab, the TKIs, where you really needed substantial expression, for those to add. But with the ADCs not being as dependent, at least in the case of the fam-trastuzumab deruxtecan, on really high levels in that being a driver. But just using the antibody as an address label to deliver the toxin and the therapeutic benefit was much more dependent on sensitivity to the toxin you're delivering rather than on shutting down HER2 signaling. When we started to see benefit for the lower levels of expression, testing came both under renewed scrutiny. Language got muddy again, with people making up new terms. So first on the field was the HER2-low category.
Badve: It reflects a different way of thinking about how the drug works. As you correctly pointed out, classic trastuzumab required internalization of the drug and the receptor and downregulation associated with that. And as you pointed out, with ADCs, internalization of the drug is no longer a necessity.
In fact, many people believe that even the presence of the protein is no longer a necessity. We have the hypoxic environment in which cathepsins of a variety of sort might be cleaving the linker, and therefore, you have a generalized toxicity in the neighborhood of the tumor rather than even being directly pointed at the tumor.
And the reason to think about it is partly because if you have one or two cells positive among the millions of cells that are there in a tumor, which technically qualifies as a HER2-low tumor, why would we see an impact of a drug? The other reasoning to think about is all breast epithelium has some numbers of HER2 molecules. Is the formalin fixation and the processing that we do destroy X number of molecules, and therefore, they become undetectable when the molecules are present below a hundred thousand or something like that molecules per cell. And if the antibody drug is sensitive enough, we may have a negative assay, yet there's enough protein for the antibody to bind the cell or be attracted to the cell and cause the damage it causes. And therefore, all these questions that now arise are what is the lowest level of protein that is required. Are there any tumors that are really HER2 negative, is zero that is detected by IHC genuinely an indication for not treating patients with ADCs, such as trastuzumab deruxtecan? Are all the ADCs going to behave in a similar manner? Because, as you know, there are at least half a dozen or more of anti-HER2 ADCs that are gradually coming onto the market, and all these questions are going to be extremely challenging to tease out what is the least amount of protein that is required. Are our assays competent enough to detect the amounts of proteins, or is the HER2 testing going to be historical? What I'm told by some clinicians, and again, I would love to hear your opinion on it, is that if the patient is classic HER2 positive — that is, 3+ or FISH positive — people would still prefer to use a non-ADC therapy, TCHP or something like that, or THP, while if you have lower levels of HER2, then they would think of ADCs as an option because of the different toxicity profile. And again, a lot of things will evolve as we understand and are better able to manage the toxicities of the drugs. I'd love to hear your thoughts on that.
Miller: I think you frame the shift in the focus of testing perfectly and in a way that I haven't heard described quite so eloquently. Initially, the focus was on identifying those with such high levels that it was a driving factor and you needed to shut down that signaling pathway with those specific targeted agents.
But as you're using them more to deliver therapeutics, it's really a question of what is the lower threshold rather than the upper threshold? And the other thing that will challenge testing and challenge those of us in the clinic is that what is a lower threshold might be drug-specific, depending on the mechanism of the drug, how stable is the linker, how much do you need truly antibody to bind to the receptor on at least some portion of the tumor cells? To see activity there? That's going to make this very challenging because I could imagine different ADCs having a different lower threshold, and the ASCO-CAP guidelines and the way IHCs are reported are going to struggle to give us that nuance or to know where that is. So I don't know that there's a simple solution. I suspect that Press would argue that doing more molecular assays with fresh tissue — so you can measure the amount of protein and not have to deal with both the false positives and false negatives that come with fixation and antigen retrieval — is the way to go. But that requires a whole change in operational systems that we have not had to do in a couple of decades.
Badve: Right, and the other thing is we have primarily focused on the antibody, and as ADCs have come along, there's an increasing focus on the linker as the major player in addition to the antibody. That linkers, the stability of the linker, the amount of destruction of the links, and the free drug that is there, and someone has even said that maybe ADCs are just a metronomic way of giving chemotherapy. Because you have the association of the drug, only 3% or 5% of it is bound to the antibody, and a lot of it is free. Don't quote me on that; this is hearsay. But the amount of free drug that exists, which is bound to proteins such as albumin and other things, and platelets, all of those could also have an important role.
There's a lot that we don't know. And as we get clarity on the role of the protein, the role of the linker, the role of the toxins, and why all of those are topoisomerase 1s can be the French data from San Antonio was very interesting about sandwich therapy vs giving it back-to-back. All those things will shed more light. As we get better clarity on patient management, we hope we will have greater clarity on testing because we have no idea what we are doing. We are essentially following the lead of the drug industry and the clinical trials.
Miller: The testing and the shift in testing and the terminology have also raised real questions about biology. And is this HER2-low group defined by testing whether you include the ultralows or not, is that group biologically distinguished in some way that is important, separate from the likelihood of response to an ADC? I've gotten frustrated at people asking questions about the HER2-low subset in clinical trials of therapies that have absolutely nothing to do with ADCs or HER2 or HER2 targeting. And other than marketing, it's not clear to me why they care. Because I don't see a biologic link there.
Have I missed something? Is that group biologically distinguishable in some other way?
Badve: Not to the best of my knowledge, but 25 years ago, when I was at Northwestern, we were looking at some tamoxifen-resistant tumors that Dr Craig Jordan had developed, and we showed that there was an HER2 overexpression in MCF-7 tumors following tamoxifen therapy. It didn't reach the level of 3+. It wasn't FISH positive, but the proteins were there, and we published that in early 2000. So these kind of data have been in existence for a long period of time, suggesting that HER2 upregulation was one of the mechanisms of resistance, and it seemed to be a relative, nonspecific mechanism, with words that are not particularly based on science but something that happens in a variety of tumors due to stress environment that is there. Now the question basically is if we know that this is a reactive mechanism and is not based in genomics, and it's somewhat a transient mechanism, we really can't separate it out as a distinct biological category. At least that's my way of thinking about it. People might have different thoughts about it, but I do not think it's permanent. It's a transient phenomenon that is occurring in response to a stress environment.
Miller: Thank you, Sunil. I always learn something when I talk to you. Today, we've talked with Dr Sunil Badve about HER2 testing and how HER2 testing has evolved over time as the need for that clinical testing has shifted. Our conversation has highlighted the critical link between the development of therapeutics and the development of medical testing, and the need for people in the clinic who are using these results to make decisions to work carefully with their pathologist, and for pathologists to understand how their reports are used to make those decisions. HER2 testing has evolved over the past two decades and will continue to evolve as therapeutics evolve, and this will continue to require this collaboration.
Thank you for tuning in. Please take a moment to download the Medscape app to listen and subscribe to this podcast series on breast cancer and HER2. This is Dr Kathy Miller for the Medscape InDiscussion podcast series.
Listen to additional seasons of this podcast.
Resources
Inhibition of Tumor Growth by a Monoclonal Antibody Reactive With an Oncogene-Encoded Tumor Antigen
Reversal of Tamoxifen Resistant Breast Cancer by Low Dose Estrogen Therapy
Medscape © 2026 WebMD, LLC
Any views expressed above are the author's own and do not necessarily reflect the views of WebMD or Medscape.
Cite this: HER2 Positive, Low, Ultralow: What Testing Means in the ADC Era - Medscape - Feb 24, 2026.
Comments