A new perspective on HER2 testing

Evolving lab practices to match clinical advances

Speaker: Prof. Dr Abeer Shaaban MBBCh, MSc, PhD, FRCPath​

5th June 2024

This non-promotional event was organized and funded by AstraZeneca and is intended for healthcare professionals only. Z4-65702. DOP: June 2024. These questions were answered by Prof. Dr Shaaban during the Q&A session of this Lab Talk. The answers reflect her professional view and expertise and do not reflect the views of AstraZeneca. This content is for educational purposes only. AstraZeneca does not guarantee that the suggestions and methods implemented here will work for your institution. Not for further distribution.

Q1: Does the pilot NEQAS scheme include FISH or is it only HER2 IHC?

It’s only IHC at the moment, but NEQAS have a separate HER2 FISH scheme. This is particularly done for HER2 low, to make sure that we as labs stain it correctly and interpret correctly.

Q2: How have you changed your magnification approach based on HER2 low?

For HER2 low by definition, you probably won’t see much at 5x or 10x.

Try to assess, go low power first and see if you can see any staining or not. If you cannot see any at 5x, this is not strong. Go to 10x and see if you can get any staining. If you can see some staining there, then there is a little bit at least of a moderate staining. Then go higher and assess the rest. The idea is to see the proportion of cells that will have complete or incomplete staining and take it from there. Generally, if you have complete staining around 10%, you’re in the 2+ category, again it could be HER2 low at the end. So, go from very low power to high power and assess the percentage and completeness and try to categorize it.

Q3: What would you highlight as the key pre-analytical variable to optimize HER2 testing, especially regarding low levels of expression?

Fixation is a key pre-analytical variable. Many of us struggle when we receive our specimens from remote areas, we don’t always have the theatres next to us and by the time we received the specimens, they are already a few hours or maybe a day later and dependent on someone being there to slice them and fix them. So, making sure that we receive the specimen as soon as possible and deal with it quickly to fix that is important, as that will lead to loss of the antigenicity if we’re not careful. You need to have this agreement with your clinical team. As I said, if the specimen cannot be transported immediately to you, or must be kept overnight, it’s better to keep it overnight in a fridge, not in formalin. And then send it in the morning to the pathology lab to be sliced and fixed immediately. It’s not good to put a specimen in formalin without slicing, which still happens. You will get very slow fixation and get fixation of the outer surface but not the centre of the tumour. We can fix this by coordinating with the clinical team.



Q4: What do you think are the key aspects to include in the reporting? For example, descriptive terminology, considering the end user and the clinical context.

For the protein expression we need to provide the final classification, is it 0,1, 2 or 3. If it is 3+, we normally add a comment that it is positive for HER2 protein overexpression and will be eligible for anti-HER2 therapy. For FISH, you can’t tell whether it is positive or negative to lead to FISH for 2+, so another comment for FISH would be on the copy number for HER2 and CEP17 and the final conclusion, what’s the ratio and what this means? Currently, you can add a comment on top of that to say that for example, if it’s 1+ plus, this is a HER2 low tumor.

A comment on what it means for the clinician is useful. For example, when you have heterogeneity, what’s the type of heterogeneity? This will need discussion. For FISH, you would comment on each heterogeneous area separately and discuss in the MDT. It is not as black or white and in some difficult cases discussion is required with the clinical team, particularly in the heterogeneous category. So, nothing has changed from what we do now, but if you want to add to it a comment on the HER2 low category, you can do that. from there. Generally, if you have complete staining around 10%, you’re in the 2+ category, again it could be HER2 low at the end. So, go from very low power to high power and assess the percentage and completeness and try to categorize it.


Q5: What would you say are the key technical points to consider when it comes to FISH to ensure accurate analysis and reporting?

For quality assurance, the fixation is important. It is usually useful as well to select areas that do not have a lot of DCIS. If the tumor is sparse or there is a lot of DCIS or surgical changes, it is useful to mark the area to be assessed, particularly if the tumor is being assessed first by the biomedical scientists or even by the pathologist, as it is difficult to assess morphology in the dark background. If the tumor is small, sparse, or with a lot of DCIS, make sure that you mark the area to be assessed. With the usual quality assurance, it is advisable to assess both CEP17 and HER2 copy number, and in the UK for example more than six for HER2 copy number and anything more than two in ratio is regarded as positive. There are some variations with the ASCO CAP guidelines in the borderline category, but broadly speaking that is the interpretation. If a case is borderline (the ratio is 2), it is recommended that we assess a larger number of cells and for a second assessor to do that. For example, you can assess 20, 20 and then do another 20 to assess 60 in total, and get two people to assess, particularly when you are at the borderline whether to call it positive or negative.

Q6: How to minimize variability in immunohistochemical analysis results in these cases depending on the antibody clone used?


The most commonly used antibodies generally in the EU are the VENTANA and the Dako HercepTest. The studies that have been done so far have shown that there is a difference, so depending on the antibodies you use the results will not be identical and that’s something to keep in mind. There are some studies that have shown that Ventana 4B5 picks up a lot more of the HER2 low category than Dako, but Dako now have got their monoclonal test that they project is picking a lot more than before. It’s important to identify which one you’re using and recognize that means they will not be the same. I think the Ventana 4B5 is the most commonly used and it is FDA approved. In the UK, that’s the most used antibody as well. However, if you’re using another one, it’s not a problem, just make sure that it works well, the quality assurance is good and that you participate in the relevant quality assurance scheme.

Q7: What is your approach to borderline or complex cases in HER2 expression, do you get a second opinion?

In general, if it is a difficult case to interpret, I would show it to one of my colleagues. Sometimes it doesn’t matter, for example, if the difference is between you calling it 3+ or 2+, if you call it 2+ and FISH it, it will be positive, so you will get the same results, but you probably will take longer. The most difficult cases I find are the differentiation between the ultra-low and the HER2 low where there is a bit of expression, but I might call it above 10% and my colleague calls it less than 10%. We are looking at the use of AI and digital pathology in this area to try to quantify the expression in a standard way because it is difficult by eyeballing to make this distinction. However, it might not be that important either, because recent announcements have shown that even the ultra-low is important. So, we might in the future conclude that we score HER2 as null, complete absent and zero and the rest are all relevant. This would make life much easier for us as pathologists because we know that we have higher concordance if we use this scheme.

Copyright © 2024, Diaceutics. All rights reserved