Challenging Cases in HER2 Testing

Course Director

Michael F. Press, MD, PhD

Michael F. Press, MD, PhD
Professor of Pathology
Norris Comprehensive Cancer Center
University of Southern California
Los Angeles, California

CME Credit
After reviewing the content below,
click here for a free, instant CME certificate.

Part 2 of a 2-part series

Dr. Press provides expert feedback to the questions submitted by your peers during a recent survey on this topic.


Accurate HER2 testing is crucial in identifying patients with breast cancer who may respond to HER2‐targeted therapy. However, pathologists face a variety of challenges with respect to tissue handling and processing, determining which testing methodology to use, and knowing when it may be appropriate to re-test. In the second segment of this two-part series, Dr. Michael F. Press discusses three challenging scenarios, submitted by your colleagues in pathology, that shed light on common controversies in the realm of HER2 testing.

Listen to This Presentation Now
Subscribe to Future Podcasts
How should a patient in the HER2-equivocal range, in other words, immunohistochemistry score of 2+ or a CISH [chromogenic in situ hybridization] ratio of 1.8 to 2.2 be treated?

Dr. Press: A HER2-equivocal result isn't very helpful to either the pathologist or especially the clinician or the patient, because they are faced with a situation where they have to either treat with a HER2-targeted therapy or not treat with a HER2-targeted therapy. So these cases need to be resolved.

One of the ways of resolving it is to send the case out to a reference laboratory that routinely deals with these kinds of issues. Oftentimes, academic laboratories do so, as do many commercial laboratories. In the end, though, this has to be addressed.

The IHC 2+ [cases], of course, can be tested by FISH [fluorescence in situ hybridization] to determine whether they are HER2-amplified or not.

The equivocal range for FISH ratios between 1.8 and 2.21 is going to be revised to the FDA-approved criteria of a ratio of 2.0—either greater than or less than—to determine the HER2 status, in the ASCO-CAP [American Society of Clinical Oncology–College of American Pathologists] guidelines that will be coming out later this year. So this equivocal range for FISH is going to be removed, as will the FISH ratio of 2.2 that has been recommended for HER2 gene amplification.

Now those cases that show aneusomy—where there is an increase in the average copy of HER2, let's say above 4 or even in the greater than 6 copies of HER2 per tumor cell nucleus, but also an increase in the chromosome 17 centromere count of a similar nature, so that the ratio winds up being less than 2—suggests that this is an example of chromosome 17 aneusomy and not HER2 gene amplification. And of course, this also will depend a little bit on the way the fluorescent signal appears in each [of the] nuclei. Usually in HER2-amplified cases, the HER2 signals are grouped, even if they are relatively small amplicons of three signals, for example. They are often relatively geographically circumscribed in the tumor cell nuclei, so that this assessment can be made of a low-end amplified case, whereas in most cases it's a chromosome 17 aneusomy. [With chromosome 17 aneusomy,] usually the green chromosome 17 centromere signals are in proximity to the orange or red HER2 signals by FISH, and they are often randomly distributed throughout the nuclei so that the HER2 FISH signals tend not to be grouped.

And in that setting, it's relatively easy and straightforward to make the distinction between aneusomy and low-end amplified cases. Our experience, which was published in 2010 demonstrated that those cases that had chromosome 17 aneusomy did not respond to HER2-targeted therapy, and we did not consider these cases to be candidates for this type of [HER2-targeted] treatment.2

Back to top

Many clinicians request retesting for HER2 in recurrent metastatic breast cancer that was initially classified as HER2-negative. Is there a justification for this approach?

Dr. Press: The short answer is yes, there is a justification for doing so at this time. Initially, in the earlier studies of HER2 as an alteration in clinical trials, of especially trastuzumab, where patients had metastatic breast cancer and were being screened for entry to these trials, the metastatic disease had not been biopsied, and so patients' primary breast cancers were used to determine the HER2 status. Those patients that had HER2-positive disease were candidates for entry into the clinical trial and treatment with trastuzumab, or also subsequently lapatinib. So the primary disease was taken as a surrogate for the metastatic disease in these phase 3 clinical trials. As a first approximation, I think that's reasonable because the data show that it's not often that HER2 status changes between primary and metastatic disease.

There are some published papers that indicate a relatively higher rate of change between primary breast cancer and metastatic breast cancer,3-6 but in many of these publications that find a higher rate, say in in the 20% range, there are also questions about the methods that are being used, especially immunohistochemistry, which may be variable in itself. The fluorescence in situ hybridization assays that have compared primary with metastatic disease demonstrate a relatively low rate of alteration or change in the HER2 status from primary to metastatic disease. Nevertheless, the ASCO-CAP guidelines that will come out later this year are going to recommend that metastases—when they are readily accessible and can be biopsied—should be biopsied for a second determination of the HER2 status, even when it's been determined initially in primary disease.

I would also mention when these patients are treated in the adjuvant setting with HER2-targeted therapies such as trastuzumab, these patients' tumors under the selective pressure of this kind of treatment could potentially, eventually lose this alteration through genetic evolution.

It may turn out that in the end one needs to be concerned in both directions, both HER2-negative disease evolving over time into HER2-amplified breast cancer, and the reverse in a treatment setting using targeted therapies. At this point, it's unresolved exactly how frequently that occurs, but it will be recommended that where metastatic disease is accessible, that the patients be re-biopsied.

Back to top

What should you do with regard to HER2 testing if the only available tumor tissue has sat in the refrigerator over a 3-day weekend without any formalin?

Dr. Press: Initially, of course, [the tissue] needs to be fixed and processed for routine evaluation by histopathology. Then I would recommend that the HER2 testing be performed by fluorescence in situ hybridization.

DNA is a more stable macromolecule than any of the others that we characterize. RNA is probably the most labile and degrades the most rapidly; protein would be next; and DNA would be the most stable.

The other advantage of using the fluorescence in situ hybridization assay to evaluate the gene is that you have built-in internal controls in the tissue sample itself. So, for example, you won't get a false negative result by FISH unless you score normal cells instead of tumor cells. There are, of course, benign cells in every breast cancer, the reactive stroma and endothelium and inflammatory cells. One can simply look at the FISH result, examine the normal cells and confirm that there is hybridization for the HER2 gene in the normal cells, and that it is at the appropriate level of two copies per normal cell, and that the ratio of HER2 to chromosome 17 is approximately equal. So one has an internal control that can confirm that the procedure worked in the tissue sample. Then one can move on to the tumor cells and evaluate whether the assay has been successful and determine the HER2 status in those samples.

Now although the ASCO-CAP guidelines from 2007 required formalin fixation from 6 to 48 hours, which should be initiated within a half hour of the patient sample becoming available,1 this is going to be modified in the upcoming guidelines to [be] from 6 to 72 hours.

The disadvantage [in the case presented] here, of course, for immunohistochemistry is that there is no built-in internal control. The half hour time to fixation has been breached, and there is no way of being certain that an immunohistochemistry negative result isn't a false negative.

On the other hand, there are some publications in the literature that show that the time to fixation is enormously forgiving in terms of getting a positive result by either immunohistochemistry or FISH. So this is still a gray area without very much data, but I would strongly recommend FISH for these cases as a way of resolving this issue.

Back to top

Please click here to redeem your CME credits now


  1. Wolff AC et al. J Clin Oncol. 2007;25:118-145.
  2. Downey L et al. Clin Cancer Res. 2010;16:1281-1288.
  3. Fabi A et al. Clin Cancer Res. 2011;17:2055-2064.
  4. Amir E et al. J Clin Oncol. 2012;30:587-592.                 
  5. Niikura N et al. J Clin Oncol. 2012;30:593-599.
  6. Chia S. J Clin Oncol. 2012;30:575-576.

The materials presented here are used with the permission of the authors and/or other sources.
These materials do not necessarily reflect the views of Answers in CME or any of its supporters.


This activity is supported by an educational grant from Genentech.

Additional support provided by Penn State College of Medicine and Answers in CME.

Privacy Policy || To receive e-mail updates, click here

Please contact with any questions, comments, or feedback about our programs.

About This Answers in CME Activity

Answers in CME, and Penn State College of Medicine are responsible for the selection of this report's topics, the preparation of editorial content, and the distribution of this report. The preparation of Answers in CME reports is supported by educational grants subject to written agreements that clearly stipulate and enforce the editorial independence of Answers in CME and Penn State College of Medicine. Our reports may contain references to unapproved products or uses of these products in certain jurisdictions. For approved prescribing information, please consult the manufacturer's product labeling. No endorsement of unapproved products or uses is made or implied by coverage of these products or uses in our reports. No responsibility is taken for errors or omissions in reports.

Hardware/Software Requirements:

For the best online experience, we recommend using the latest versions of these supported browsers: Google Chrome, Microsoft Internet Explorer, Mozilla Firefox, Safari.
Depending on your browser of choice, additional software, such as iTunes and Adobe® Reader may be required.

Answers to Pathologists' Questions: Changing Perspectives on Optimal Determination of HER2 Status in Breast Cancer

  1. Best Practices in HER2 Testing for Breast Cancer