Expert Sheds Light on Issues With ASCO-CAP HER2 Guidelines

Michael F. Press, MD, PhD

Due to discrepancies in HER2-testing results, ASCO and the College of American Pathologists (ASCO-CAP) recently released new guidelines for the use of fluorescence in situ hybridization (FISH) testing for HER2 in breast cancer.

The guidelines separate in situ hybridization (ISH) into 5 groups based on HER2 FISH ratio and average HER2 gene copy number per tumor cell. According to the ASCO-CAP guidelines, breast cancers in groups 1, 2, and 3 are interpreted as ISH-positive, group 4 as ISH-equivocal, and group 5 as ISH-negative. These classifications differ from the original FDA-approved criteria for HER2 gene amplification.

Using data from 3 Breast Cancer International Research Group studies, researchers at the University of Southern California conducted a retrospective study comparing the original FDA-approved criteria for HER2 gene amplification with the current ASCO-CAP guidelines. Their findings support the original categorizations of HER2 by FISH status.

“The most significant thing we learned is that compared to the original FDA-guidelines, we did not find an advantage to the current ASCO-CAP guidelines in predicting correlations, either with HER2 protein expression or with patient outcomes in terms of overall and disease-free survival,” said lead study author Michael F. Press, MD, PhD, a professor in the Department of Pathology and co-leader of the Women’s Cancers Program at Norris Comprehensive Cancer Center. “The FDA approach is perfectly adequate and we do not have reason in most cases to diverge.”

OncLive: How was the study conducted?

To learn more about the study, OncLive spoke with Press about the trial design and exactly how the 2 sets of guidelines differ from one another.Press: We looked at samples from over 10,000 women from 3 clinical trials and stratified them according to HER2 FISH status as either HER2-amplified or HER2 nonamplified. Current ASCO-CAP guidelines include a category that they call “HER2-equivocal,” which is HER2 unknown.

In our original screening using the FDA-approved guidelines, they were sorted into 2 different groups, amplified or nonamplified. Since we had these data on HER2 by FISH and we also have the data of HER2 by immunohistochemistry (IHC), we could re-sort these cases according to the current ASCO-CAP FISH guidelines, which includes 5 different groups that they have categorized.

These groups are then determined to be ISH-positive, ISH-negative, or ISH-equivocal. They have 3 categories and we usually only use 2. Therefore, we validated with FISH and sorted our cases into the 5 categories that they have, calling them groups 1 through 5, and then analyzed the data that was downstream of those patients, in terms of their clinical outcome and in terms of the tissue samples—how they were categorized by IHC. Our findings were that IHC was more strongly correlated with the original HER2-gene amplification status determined by the FDA guidelines rather than the ASCO-CAP guidelines.

What are the biggest differences between the original FDA-approved guidelines and the ASCO-CAP guidelines?

Having said that, it is important to keep in mind that there is about a 95% agreement rate between breast cancers analyzed by either of these approaches. The differences are in about 5% of women. Therefore, it is not a huge difference, and the differences are in groups 2, 3, and 4, which are basically the groups created by the ASCO-CAP guidelines that are not part of the original FDA guidelines.The first big difference was seen in group 2. To understand the differences, you need to understand that the number of HER2 gene copies is represented as an average, and that is compared to a control that is on the same chromosome, ideally, as the HER2 gene. The standard approach is to use chromosome 17 centromere, which is an approach that has been used since about 1991. When the ratio between HER2 and chromosome 17 centromere is greater than 2, it is referred to as HER2 gene amplification.

What the ASCO-CAP guidelines did is break this up into 2 groups: one group where the average HER2 copy number is greater than or equal to 4 and one in which it is less than 4. HER2 gene amplification represents a substantial increase in the number of copies of the HER2 gene. Four is not a substantial increase, but 20 or 30 are substantial increases.

Our previous data that we published in 1997 suggested that the behavior of breast cancers in women that did not have at least 4 copies of HER2 was relatively less aggressive, like that of women who had conventional HER2 nonamplified breast cancer. Since 1997 we have, in our consultation practice, called attention to these patients that are so-called “HER2-positive” patients with a ratio rate of HER2 but probably not gene amplification.

What others and we have suggested is that this group represents the loss of the denominator. In other words, this is something that has happened to the chromosome 17 centromere. Therefore, the ratio is greater than 2, but not because the HER2 gene has increased copy number substantially, but because the internal control is lost. This represents only 1% of breast cancer, so it is a small group, but it is a group that we have handled differently than the current ASCO-CAP guidelines recommend.

The guidelines consider them to be ISH-positive, and we do not consider them positive and have not for a long time. Many of these patients were accrued to a clinical trial for trastuzumab (Herceptin) in the adjuvant setting. In that clinical trial, about half of the patients received trastuzumab and the patients who received trastuzumab didn’t show a benefit from getting the drug, with a hazard ratio of approximately 1. The clinical outcomes of those who did and did not receive trastuzumab were comparable, so we would argue that based on IHC and response to trastuzumab group 2 does not appear to behave like HER2-amplified disease.

The other group that was different was group 3. Group 3 is a group where that ratio is less than 2 and the average HER2 copy number is greater than or equal to 6. This is a group of cases that the ASCO guidelines require the laboratory to store as being HER2 ISH-positive or amplified. We have not interpreted these cases in that way. In the past, we have interpreted them of being composed of 2 different groups of breast cancers; one that is HER2 amplified and another subgroup that is HER2 nonamplified. We describe that in both the JCO paper and also in the Archives of Pathology & Laboratory Medicine paper that is published. We describe how one can characterize these, and we find that those that have been called amplified are only about 1 out of every 5 cases; however, they have high levels of HER2 protein expression and HER2 gene copy number, on average.

In these cases, we consider to have extended the area of chromosome 17; we consider it to be rather large and extended so that it includes our control gene chromosome 17 centromere. This is so that both the numerator and denominator of that ratio are increased. Typically, they have 10 or copies of HER2 and a similar number of chromosome 17 centromere. They have a very characteristic appearance when we look at them in the microscope, as well. We feel that the ratio is less than 2 because the denominator is artificially increased. We think of this as not a single subgroup of patients, but a category that has at least 2 different classifications of gene amplified and non-amplified. Group 3 also accounts for a little less than 1% of breast cancers.

The final group that we saw a difference in was group 4. Group 4 are the cases that have a ratio of less than 2 and an average copy number between 4 and 6. These are the cases that the current guidelines require that they should be considered equivocal. They represent somewhere in the range of 4% to 6% of breast cancers.

What impact will these findings have?

In our study, they were in the 4% to 5% range. In these cases, they do not show HER2 overexpression by IHC and these patients are accrued to 1 of our 3 clinical trials and their outcome in those trials are statistically indistinguishable from the HER2 nonamplified breast cancers that are referred to as group 5. Based on protein expression and overall survival, we consider these patients to have breast cancers that are HER2 nonamplified.It is pretty clear that, at our laboratory, we believe in our data. We will be using the findings of this study and another study published in the Archives of Pathology & Laboratory Medicine a few months ago. Between those 2 studies, they represent nearly 18,000 women in terms of determination of the HER2 status by FISH and IHC. That is a substantial amount of data and that is the approach that we will use. We have not really been using the ASCO-CAP guidelines and we will not be. The data that we have generated here suggests that this is a reasonable strategy.

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.