Charna Albert
June 2025—The authors of a recently published study suggest repeating HER2 testing on the excision specimen for the small number of breast cancer biopsies with group two and group four FISH results.
The aim of their study was to determine if FISH group two and group four cases change HER2 status after repeated testing on additional specimens (Klaric KA, et al. Arch Pathol Lab Med. Published online March 11, 2025. doi:10.5858/arpa.2024-0305-OA).
“The current guidelines say you may consider repeating on excision,” says study coauthor Phillip Williams, MD, MSc, staff pathologist at the Ottawa Hospital and Eastern Ontario Regional Laboratory Association and associate professor at the University of Ottawa. “When we looked at our own practice in our hospital, some were repeated and some were not.”
Groups two and four are considered HER2 negative by the 2018 ASCO/CAP guideline for HER2 testing in breast cancer, unless concurrent immunohistochemistry is 3+. “They’re rare cases,” Dr. Williams says. “There’s not much data out there regarding these cases. But because Herceptin is such a powerful drug with such great outcomes, it’s important we identify any potential patients that may respond.” And repeat testing of a small number was unlikely to burden the laboratory. “We decided to look at our institution, see how many of these cases we had,” and determine if HER2 status would change with the excisional testing, making patients eligible for treatment.
He and his coauthors identified all HER2 FISH cases reported at the Ottawa Hospital from 2016 to 2023 and the results of repeated testing. To his surprise, more than 10 percent of the group two cases and nearly 20 percent of the group four cases became HER2 positive. “The results were unexpected,” Dr. Williams says. “I expected the testing on excision to be more in line with overall HER2 results—that HER2 testing is 98, 99 percent concordant between biopsy and excision.” Thanks to their findings, he says, the laboratory now repeats HER2 testing on excisional specimens in all such cases, instead of at the pathologists’ discretion.
With the 2018 ASCO/CAP guideline update, groups two and four were deemed HER2 negative unless concurrent immunohistochemistry is 3+, and these designations were affirmed in the 2023 guideline update. Previously, they had been designated positive and equivocal, respectively. “Equivocal was neither positive nor negative, and how to address those was a little unclear previously,” Dr. Williams says. “Before, the guidelines said you should repeat it [HER2 testing] on the same specimen, potentially with different probes,” and that led to multiple retests with different results. “The new guidelines were nice in that they clarified it’s a negative result,” but noted studies that found “sometimes they become positive” on repeated testing, he says. The guidelines say that repeated testing of the excisional specimen from the same patient may be appropriate in the setting of groups two and four.
The ISH results for these groups represented a small proportion of cases in the initial clinical trials for trastuzumab. Further investigations and retrospective review failed to demonstrate benefit over chemotherapy alone for these groups, although in the case of group two the sample size was too small to draw definitive conclusions.
Wang, et al., studied repeat HER2 tests at their institution, performed for cases with available alternative tumor samples and compared with initial testing following the 2018 ASCO/CAP guideline (Wang M, et al. Clin Breast Cancer. 2023;23[4]:415–422).
Only one of 23 group two cases was HER2 positive (zero of 18 in primary and one of five in metastatic/recurrent tumors). Of 13 primary tumors with repeat HER2 results, 10 remained HER2 negative. Three changed to HER2 positive (group one and IHC 2+).
In their own study, Dr. Williams and colleagues identified 5,695 cases with HER2 FISH testing: 101 of group two (1.8 percent) and 194 of group four (3.4 percent). Of the group two cases, initial HER2 FISH testing was carried out on 90 primary breast biopsy specimens (89.1 percent) and 11 biopsies from metastatic disease sites (10.9 percent). The tumors were predominantly grade two (n = 58, 57.4 percent) and ER positive (n = 79, 78.2 percent). The group four cases were from 165 primary breast biopsy specimens (85.1 percent), 17 primary breast resections (8.8 percent), and 12 metastatic biopsies (6.2 percent). Like group two, the group four cases were mostly grade two tumors (n = 101, 52.1 percent) that were ER positive (n = 159, 82 percent). At the Ottawa Hospital, first-line HER2 testing is performed with IHC, and for cases where the HER2 is equivocal (2+), FISH is then performed. As such, all initial group two and four cases were HER2 IHC 2+.
“I know a lot of institutions go to FISH testing first,” Dr. Williams says. But for group two or four FISH results, “it is recommended that you either look at the immunohistochemistry concurrently—which we do first—or you have to go back and do it anyway. We do it up front and then we already have the results.”
Of the 90 initial group two breast biopsies, 42 (46.7 percent) had HER2 testing repeated on the surgical resection specimen and were included in the study. Five (11.9 percent) of the excisional specimens were HER2 positive by FISH. All five cases showed HER2 FISH group one results (HER2:CEP17 ratio ≥ 2.0 and average HER2 copy number signals ≥ 4.0 per cell). None demonstrated overexpression by IHC (all were IHC 2+). With more than 10 percent of patients becoming positive, Dr. Williams says, “it’s worthwhile repeating these, especially as they’re rare.”
Of the group four cases, 110 of the 165 initial breast biopsies (66.7 percent) had repeated HER2 testing performed on the excision specimen. Nineteen (17.3 percent) became HER2 positive: 17 of these were group one on HER2 FISH, with the other two classified as group three (HER2:CEP17 ratio < 2.0 and average HER2 copy number signals ≥ 6.0 per cell). None demonstrated overexpression by IHC. One case was HER2 IHC negative (1+), but there was amplification by FISH testing (HER2:CEP17 ratio = 3.2; average HER2 copy number = 5.8). Although a negative IHC would typically preclude further testing by FISH, “due to the previous FISH results we thought it would be better to retest,” Dr. Williams says. And in that case, he recalls, the only residual carcinoma present was in a lymph node, suggesting inadequate fixation may have resulted in the negative IHC.
For the resection specimen, Dr. Williams and his coauthors suggest performing direct FISH or concurrent IHC/FISH. If they’re a group two or four on biopsy, “our results have shown they’re more likely to be a positive result on excision, and the vast majority of our cases were 2+ on the excisional specimen,” Dr. Williams says. “But in cases where the fixation time or the ischemic time is unknown or extended, it may be a good idea to go direct to FISH, irrespective of the results on immunohistochemistry.”
Another surprise awaited Dr. Williams and his colleagues as they reviewed the hospital’s breast biomarker synoptic reports.
HER2 FISH testing, he and his coauthors note, uses probes targeting the HER2 gene and the centromere of chromosome 17. The HER2:CEP17 ratio and average number of HER2 signals per cell are used to determine the HER2 category. They had expected the HER2:CEP17 ratio and average HER2 copy number signals on initial biopsy to predict HER2-positive status on repeated testing of the excision specimen.
In practice, Dr. Williams says, “We may look at cases that are close to the threshold and then preferentially choose those to retest.” In other words, a pathologist might select for repeat testing a group two or four case with a HER2:CEP17 ratio and average number of HER2 signals per cell closer to that of a group one or group three case (both considered HER2 positive). But in their review, the HER2:CEP17 ratio and HER2 copy number signals on initial biopsy did not appear to predict HER2-positive status on repeated testing of the excision specimen.
“We thought it would be predictive. When you have something close to the threshold, you would think that case would be more likely to be positive on repeat testing,” Dr. Williams says. “But in our study, it didn’t seem to affect anything. That’s why we recommend they all be retested, because it’s difficult to predict which will be positive.”
A tumor is considered HER2 group two by FISH when it has a HER2:CEP17 ratio of two or higher and average HER2 copy number signals less than four per cell. A group four tumor has a HER2:CEP17 ratio of less than two and average HER2 copy number signals of four or higher but less than six per cell. For the five group two cases that became HER2 positive, the HER2:CEP17 ratio on initial biopsy ranged from 2.0 to 3.0 (mean, 2.3). For the cases that remained HER2 negative, the initial biopsy ratios ranged between 2.0 and 3.5 (mean, 2.2). It did look as though the group two biopsies with a slightly higher range of average HER2 copy number (3.2–3.9) were HER2 positive on excision, but the mean value (3.4) was the same as the cases that remained negative on excision (mean, 3.4; range, 2.4–3.9). “Interpretation here is also limited by small sample size,” Dr. Williams and his coauthors note.
Similarly, no appreciable difference was seen in the group four biopsies. For the 19 initial samples that became HER2 positive, the HER2:CEP17 ratios ranged from 1.2 to 1.9 (mean, 1.6) with HER2 copy numbers between 4 and 5.4 (mean, 4.6). For the cases that remained HER2 negative, the ratios ranged from 1.0 to 1.9 (mean, 1.6) with HER2 copy numbers between 4.0 and 5.8 (mean, 4.5).
Does this finding say something about the methods used to determine the HER2 category?
“Overall, HER2 testing is concordant in almost every study I’ve read,” Dr. Williams says. At the Ottawa Hospital, a 2019 quality assurance assessment incorporating IHC and FISH results found a 97 percent concordance rate in 193 cases between the initial HER2-negative breast biopsy and subsequent resection specimen. Groups two and four, however, were excluded (the assessment included only IHC 0, IHC 1+, or IHC 2+/not amplified). “It would be interesting if you teased out these group two and four [cases] and looked at those, versus repeating the other categories, to see if there’s a change,” he says.
Neoadjuvant chemotherapy, another variable they studied, didn’t appear to affect HER2 status. In group two, of the 42 cases with available HER2 results on the excision specimen, four (9.5 percent) underwent neoadjuvant chemotherapy: Two were triple-negative breast cancer and the other two were hormone positive (ER >10 percent). Only one of these four cases was HER2 positive on repeated testing of the resection specimen (ER/PR positive on biopsy). Of the 110 group four cases, 12 (10.9 percent) underwent neoadjuvant treatment, four of which were triple-negative breast cancer. On repeated testing of the excised tumor, two of the 12 became HER2 positive (both ER/PR positive on biopsy). When compared with the neoadjuvant cases in which HER2 remained negative, no association was observed between HER2 status on repeated testing and receipt of neoadjuvant therapy in both group two and group four (P = .41 and P > .99, respectively).
“I wouldn’t say it was unexpected,” Dr. Williams says of the finding. “But you would think if neoadjuvant chemotherapy is given, it may select for a clone that may be more aggressive, which HER2 usually is. We thought it possible that the cases undergoing neoadjuvant chemotherapy may be enriched in HER2 clones, but our research didn’t show that.” Triple-negative status and high tumor grade were likewise not found to be associated with HER2 status on repeated testing. “Therefore,” Dr. Williams and his coauthors write, “the recommendation to repeat testing is irrespective of these features.”
Dr. Williams and his coauthors say the change in HER2 status between initial biopsy and resection specimen may be attributable to intratumoral heterogeneity.
“This coexistence of tumor cells with varying degrees of HER2 expression and/or amplification is well described and has been reported in up to 40% of tumors with a predominance in equivocal cases,” they write, noting again that in their study, all initial group two and four results on biopsy were IHC 2+ and that most were IHC 2+ on repeated testing. (Of the group two cases, 31 were IHC 2+, 10 were 1+, and one was IHC 0; in group four, 84 were 2+, 19 were 1+, and six were IHC 0.)
Wang, et al., in their 2023 study, had similar group two case results, Dr. Williams and coauthors say, but they cite an earlier study that found that groups two and four were strongly correlated with immunohistochemistry 0/1+ status (Press MF, et al. J Clin Oncol. 2016;34[29]:3518–3528). “This finding,” Dr. Williams and his coauthors write, “raises the possibility that our institutional practice of first-line IHC assesses a subset of group 2 and 4 tumors with more borderline positive features, and that we better capture the HER2-positive clonal population in repeating the HER2 on the excision.”
Dr. Williams acknowledges that laboratory protocols could affect results. “The biopsies are well fixed,” he says. “The excisions can be variable depending on the lab, so ischemic and fixation times can be variable,” and HER2 IHC is sensitive to preanalytical variables. Results also depend on the antibody clones used.
A future direction would be to explore the rate of change in the HER2 status of these two groups in cases where the level of HER2 expression is low or negative. With repeat testing at the Ottawa Hospital formerly the pathologists’ call, Dr. Williams says, “there may have been some selection of which cases to repeat that we didn’t see.” Now that repeat testing is institutional policy, “we’re going to look retrospectively at whether repeating all of them changes the results we found—whether there was some sort of bias or unexpected enrichment by pathologists in choosing which cases to repeat.”
Dr. Williams sounds a note of caution about the study’s limited sample size. “We had over 5,000 FISH cases in our institution, but we still had only a rare number of these cases,” he says. And for group two in particular, “the efficacy of Herceptin in these cases is still unknown.” There’s also the question of how the HER2-low designation may change which patients are selected for treatment.
For now, “It would be interesting for multiple institutions to repeat these findings to see if they are concordant,” he says. “I imagine they are.”
Charna Albert is CAP TODAY associate contributing editor.