Patients Liable to Be Misclassified Under New HER2 Gene Testing Guidelines, Expert Says

Michael F. Press, MD, PhD

Michael F. Press, MD, PhD

According to Michael F. Press, MD, PhD, new guidelines for the way patients with breast cancer should be classified are a concern, as a small subset of patients are liable to be misclassified and receive the wrong treatment. Press, a member of the College of American Pathologists (CAP), has recommended changes to updated ASCO-CAP guidelines on assessment ofHER2amplification by fluorescence in situ hybridization (FISH).

Press, the Harold E. Lee Chair for Cancer Research in the Department of Pathology, Norris Comprehensive Cancer Center, University of Southern California, discussed the new ASCO-CAP guidelines for HER2 FISH testing at the 2017 Miami Breast Cancer Conference. Although the updates have resulted in no changes for approximately 90% to 95% of cases, they will result in potential disagreements for approximately 5% of HER2 breast cancer cases, according to Press.

Targeted Oncology:What is controversial about theHER2oncogene right now in breast cancer?

Press:

The guidelines were published in 2013 and 2014 and did not recommend many changes related to testing for the HER2 protein using immunohistochemistry [IHC]. Some minor revisions were made, but they didn’t substantially impact the way in which HER2 is evaluated by IHC. On the other hand, the ASCO-CAP guidelines committee did make a number of changes in the way theHER2gene is evaluated by FISH for gene amplification. These changes have created confusion about whether or not some FISH subgroups areHER2amplified, known as in situ hybridization (ISH)-positive, orHER2not amplified, known as ISH-negative. This confusion impacts approximately 5% of women with breast cancer.

Approximately 20% to 25% of breast cancers in women have theHER2gene alteration, also known as gene amplification. This means that the cancer cells have increased the number of copies of theHER2gene in each tumor cell. Whereas normal cells have 2 copies of theHER2gene, cancer cells often have greater than 20 copies per tumor cell. This leads to a great increase in such things as cell migration, cell invasion, and increased cell division. It stimulates the production of new blood vessels, and these cancer cells are much more aggressive than cancers that lack this particular alteration.

The good news is that there are several different HER2-targeted therapies that interfere with the actions of the overexpressed protein on the cell surface. And one class of these would be monoclonal antibodies that are directed against the extracellular domain, such as trastuzumab (Herceptin) and pertuzumab (Perjeta). There are also antibody—drug conjugates (ADCs). There’s T-DM1 (trastuzumab emtansine; Kadcyla), which is linked to a biological toxin, so that the drug is directed to tumor cells that overexpress HER2. When the membrane is internalized, the poison that’s linked to trastuzumab damages and destroys the tumor cells. Another class of agents is anti-HER2 small molecule inhibitors. Either one of these classes of drugs prevent the continuous stimulation of HER2 in tumor cells.

Because there are drugs that are aimed atHER2gene amplification and overexpression, it’s important to know who among patients with breast cancers has this alteration and who doesn’t. In the absence of gene amplification, there is no HER2 overexpression. The ASCO-CAP guidelines for HER2 testing by FISH have created new definitions of what is HER2-positive and what is HER2-negative. And I think these new definitions are problematic.

Targeted Oncology:How are these definitions changing, and what makes them so problematic?

Press:

In the past, women were enrolled in clinical trials of HER2-targeted drugs if they had HER2-positive breast cancers. The definition of HER2-positive by FISH in the past referred to a ratio between the number of copies of theHER2gene, compared with a control gene on the same chromosome arm that was greater than 2. The centromere is the usual control gene for calculation of this ratio. When the ratio of theHER2gene compared with the centromere (CEP17) is greater than 2, these breast cancers are referred to asHER2-amplified.

The current ASCO-CAP guidelines for HER2 testing by FISH require consideration of both theHER2-to-CEP17 ratio and the averageHER2gene copy number per tumor cell. Because the clinical trials used for FDA approval of HER2-targeted therapies did not use this definition, it was not clear how the current ASCO-CAP guidelines definitions correspond to the previous definition ofHER2amplification used for enrollment in the clinical trials.

In order to address this issue, we reanalyzed theHER2gene amplification status of more than 10,000 women previously enrolled in 3 Breast Cancer International Research Group clinical trials using the current ASCO-CAP guidelines. We found that the HER2 status for approximately 95% of these women’s breast cancers was not changed by the new guidelines. However, the HER2 status of approximately 5% of these women was different than the status originally assigned to their cancers.

Our study was published last year in theJournal of Clinical Oncology, and it reported that difference. Although this change in HER2 FISH scoring affects only a minority of the breast cancer population, these 5% of cases constitute 3 of 5 FISH scoring categories created by the new ASCO-CAP guidelines. Our findings indicate all 3 of these subgroups are mislabeled by the guidelines and should be revised.

Targeted Oncology:Do oncologists in this space share your dissatisfaction with the newer guidelines?

Press:

Many medical oncologists treating patients with breast cancer are dismayed about the uncertainty the current guidelines create for some of their patients. The ASCO-CAP guidelines FISH group that has created the most confusion is the group classified as ISH-equivocal. These women represent about 4% of all breast cancers. Because a HER2-targeted therapy “treat” or “not treat” decision must be made, assignment to an equivocal category is not acceptable for either the patient or her oncologist. Our findings indicate that the breast cancers in this equivocal category are not amplified and women with these cancers have a clinical outcome similar to that of other women whose cancers are considered ISH-negative orHER2not amplified. Some of these women with ASCO-CAP equivocal breast cancers are, nevertheless, currently receiving HER2-targeted therapy.

The equivocal group has represented a problem that comes up in about 1 out of every 20 patients. For medical oncologists treating patients frequently, this issue arises with great regularity.

The additional problem with this is that the ASCO-CAP guidelines recommend that these patients’ breast cancers can be analyzed with alternative control genes. As I explained before, there’s a numerator ofHER2copy number divided by the chromosome 17 centromere copy number. That ratio in this equivocal category is less than 2. But they [ASCO-CAP guidelines] have also permitted, or even suggested, that alternative genes on the same chromosome could be used in that denominator as comparison controls.

This chromosome has 2 arms; theHER2gene is located on the long arm of chromosome 17, the q arm. One could choose other genes that are on that arm for comparison. Retinoic acid receptor alpha is a frequently used gene on that arm. The opposite arm, the p arm, which is the short arm of chromosome 17, has a number of other genes that are used, including theTP53gene, and the Smith-Magenis Syndrome region—which is a birth defect in children that’s on that arm—and there are a couple of other genes that are used and recommended, for example, by the Mayo Clinic and the Cleveland Clinic, as alternative control genes. Many large laboratories are using all of these controls, and when any one of them produces a ratio greater than 2, they’re recommending that these patients be considered HER2-positive by ISH. Our review in our lab would be, and has been for a long time, that this is not warranted, because many of those genes on the p arm of the chromosome are deleted.

Breast cancers can increase gene copy number, or they can reduce gene copy number. On the p arm, I mentioned thatTP53is a gene that’s present on that arm, and it’s mutated in about 20% of human breast cancers. In that setting where it’s mutated, the normal copy of TP53 is frequently lost or deleted. So that whole arm of chromosome 17 has relatively frequent deletions of portions of that arm. If you have, say, a numerator that’s just a normal value, and you lose one of the 2 comparator genes that’s in the denominator, suddenly your ratio, instead of being, say, 1, goes to 2, simply because half of your denominator is lost. It doesn’t really reflect that the HER2 gene has increased in copy number. There’s a loss of the comparison gene, the control gene.

In that setting, a lot of large commercial laboratories and large institutional laboratories are reporting a ratio greater than 2 and suggesting that these patients are HER2-positive. The patients are then being subjected to a course of chemotherapy and trastuzumab that is consistent with HER2-positive disease. We’re suggesting that this is inappropriate, and the data we have on clinical outcomes in our paper also suggest that this is not warranted.

Reference:

Press MF, Sauter G, Buyse M, et al. HER2 gene amplification testing by fluorescent in situ hybridization (FISH): comparison of the ASCO-College of American Pathologists guidelines with FISH scores used for enrollment in Breast Cancer International Research Group clinical trials [published online August 29, 2016].J Clin Oncol. doi:10.1200/ JCO.2016.66.6693.