Clinicopathologic Features of 2018 American Society of Clinical Oncology/College of American Pathologists Fluorescence In Situ Hybridization Group 3 Breast Carcinoma (Human Epidermal Growth Factor Receptor 2 Chromosome 17 Centromere Ratio <2.0 and Average Human Epidermal Growth Factor Receptor 2 Copy Number ≥6.0)

Breast cancer cases submitted to our reference laboratory for HER2 FISH testing and all in-house patients diagnosed with primary, recurrent, or metastatic breast cancer tested with HER2 FISH between January 2019 and June 2022 were included. All cases with HER2:CEP17 ratio of less than 2.0 and average HER2 copy number 6.0 or greater were identified. Clinicopathologic information was obtained from outside pathology reports in reference laboratory cases and from electronic medical records for in-house patients. Specimen types included cytology, biopsies (core and punch biopsies), and resections (lumpectomies, mastectomies). The retrieved clinicopathologic information included patient age, tumor size, tumor grade, hormone receptor status, histologic type, and lymph node status. HER2 FISH results, including average HER2, CEP17 copy numbers, HER2:CEP17 ratio in the original, and, when performed, in blinded second count and HER2 IHC results, were recorded. For in-house cases, electronic medical records were reviewed for adjuvant therapy and neoadjuvant therapy (NAT), type of therapy (chemotherapy, targeted therapy, hormonal therapy), and clinical follow-up.

For reference laboratory cases received for FISH testing, we did not attempt to determine if prior HER2 IHC was performed or review those tested at the referring institution. In our institution, all breast cancer cases were tested by HER2 FISH as the primary testing method during the study period. HER2 dual-probe FISH assay was performed on 4- to 5-µm-thick formalin-fixed, paraffin-embedded tissue using Dako HER2 IQFISH PharmaDx DNA Probe Kit (Agilent Technologies, Santa Clara, California) as per manufacturer’s instructions. Each FISH slide was manually screened and counted by a trained technologist and then reviewed by a board-certified anatomic pathologist with expertise in solid tumor FISH. HER2 FISH results were classified into 5 categories as defined by the 2018 ASCO/CAP guideline.⁵  HER2 IHC was performed in our laboratory using sections from the same tissue sample used for FISH if FISH testing results were groups 2 through 4. For HER2 IHC we use the Dako HercepTest IHC Kit (Agilent Technologies). HER2 IHC results were reviewed manually by board-certified breast-subspecialty pathologists. HER2 IHC staining intensity (weak, moderate, strong) and percentage circumferential staining were, recorded, and results scored as negative (0, 1+), equivocal (2+), or positive (3+) in accordance with the guideline.⁵  For equivocal (2+) IHC results, areas of tumor with perceptibly more intense staining or more complete membranous staining were annotated to guide targeted FISH rescoring. Rescoring was done by both a second/blinded technologist and a pathologist by counting 40 additional cells.

The study was approved and conducted in compliance with our institutional review board (#00091019).

Differences between groups of categoric variables were compared using χ² and Fisher exact tests. Student t test was used to compare differences in continuous variables. P values of <.05 were considered statistically significant.

During this study period, 8800 (1204 in-house; 7596 reference laboratory) cases had HER2 FISH tests performed. A total of 142 of 8800 FISH results (1.6%) had a HER2:CEP17 ratio of less than 2.0 and an average HER2 copy number of 6.0 or greater, qualifying for group 3 following the ASCO/CAP 2018 guideline.⁵  All patients were women, and 18 (12.7%) were 50 years of age or younger. Ninety-nine specimens were from biopsy or fine-needle aspiration cell blocks, 40 were from excisions, and 3 were of unknown specimen type. One hundred ten were invasive ductal carcinomas, 7 were invasive lobular carcinomas, 10 were other histologic types, and 15 were unknown type. After exclusion of cases (n = 31) with unknown grade, 47 of 111 cases (42.3%) were grade 3. Of cases with known hormone receptor status, 103 of 119 (86.6%) were hormone receptor positive (estrogen receptor [ER] and/or progesterone receptor [PR] positive). Of the cases known to be from excision, 24 of 37 (64.9%) had T1 (≤2 cm) and 19 of 34 (55.9%) had negative lymph nodes (N0). There was no significant association between the final HER2 status (positive versus negative) of the tumor and patient age, tumor size, tumor grade, or hormone receptor or lymph node status (Table 1).

All 142 patients with ASCO/CAP FISH group 3 results (HER2:CEP17 ratio of less than 2.0 and average HER2 copy number ≥6.0) had HER2 IHC testing using sections from the same tissue sample used for FISH. Fifty-two cases (36.6%) showed negative (IHC 0 or IHC 1+) staining pattern (Figures 1, A through J, and 2). These cases were finalized as HER2 negative with a comment indicating that “in the absence of protein overexpression, it is uncertain whether patients with HER2 ratio less than 2.0 benefit from HER2-targeted therapy.” Eighty-six cases (60.6%) had equivocal (2+) IHC. A second observer blinded to the previous FISH results recounted a minimum of 40 cells corresponding to 2+ staining areas in an annotated IHC slide. Fifty-six of the 86 cases (65%) had an HER2:CEP17 ratio of less than 2.0 and an average HER2 copy number of 6.0 or greater in the targeted area, and 14 of 86 (16%) had an HER2:CEP17 ratio of 2.0 or greater and an average HER2 copy number of 4.0 or greater, leading to 70 of the 86 targeted FISH recounts (81%) being classified as HER2 positive. Sixteen of 86 targeted counts (19%) had HER2:CEP17 ratio of less than 2.0 and average HER2 copy number less than 6.0 and were classified as HER2 negative with the above-mentioned comment. The median HER2 copy number per cell was higher in cases that were classified as HER2 positive after targeted blinded recount (ie, HER2 copy number ≥6.0) compared with those that were HER2 negative (HER2 copy number <6.0) (Figure 3). Following IHC testing, 4 of the 142 cases with group 3 results showed 3+ staining pattern and were classified as HER2 positive without additional/blinded second FISH recounting. In total, 74 of 142 original group 3 HER2 FISH breast cancer cases (52.1%) were classified as HER2 positive after additional workup. There was no correlation between intensity of HER2 IHC (weak versus moderate) in IHC equivocal/2+ cases that required second/blinded FISH counting and final HER2 status (data not shown). Original group 3 FISH cases that had final HER2-positive results had significantly higher (mean, 7.2; range, 6.0–9.4) average HER2 copy numbers on original counts compared with those that had final HER2-negative results (mean, 6.3; range, 6.0–7.2) (Table 2). Average HER2 copy numbers of the 2+ IHC-guided second FISH counts were also significantly higher in HER2-positive cases compared with HER2-negative cases (Table 2).

Nine in-house cases had treatment and follow-up (average, 16.9 months; range, 1–62 months) information. Five of 7 patients who were HER2 positive received NAT, of whom only 1 (20%) achieved pathologic complete response (pCR) (Table 3). Seven of 9 (6 of 7 HER2-positive and 1 of 2 HER2-negative) patients with follow-up information were alive with no evidence of disease (Table 3).

In this retrospective study from a single institution including reference laboratory and in-house breast cancer cases, cases with HER2:CEP17 ratio of less than 2.0 and average HER2 copy number 6.0 or greater, classified as group 3 following 2018 ASCO/CAP guideline, were rare, constituting 1.6% of all FISH test results. After integrated review of FISH and IHC, a total of 74 cases (52.1%) were classified as HER2 positive. We did not find any correlation with patient age, hormone receptor status, grade, tumor size, or lymph node status of the tumor and final HER2 positivity after the workup of HER2 FISH group 3 cases. However, average HER2 copy number per cell was significantly higher in patients who were found to be HER2 positive compared with those who were HER2 negative after the workup.

In this cohort, 4 of 142 group 3 cases (2.8%) had IHC 3+ staining, a rate that is not significantly different from the 4.7% reported from another national reference laboratory.⁸  In our institution, tumors with heterogeneous 2+ or greater staining are annotated for FISH recounting. In tumors with intense circumferential staining where the staining intensity is heterogeneous, a second blinded FISH count was performed in the annotated area (rather than using the minimum 10% threshold to classify a case as 3+/positive). Studies have shown group 3 to be a heterogeneous set of patients. In one of those studies⁶  involving 3 academic medical centers including 63 patients, 31.7% were HER2 3+, 55% HER2 2+, and 13.7% HER2 0 or 1+. Results from another academic center that receives reference cases⁹  reported on 48 patients and found only 8.3% to be IHC 3+, 14.6% HER2 2+, and 77% HER2 0 or 1+. In that study, higher HER2:CEP17 ratio correlated with IHC 2+ or 3+ in a majority of the cases. Reanalysis of the HERA trial¹⁰  identified a small subset (21 of 6018; 0.3%) of cases that were originally considered negative because of a less than 2.0 ratio and 6.0 or greater HER2 signals per cell. Sixty-six percent of the cases showing HER2 copy number 6.0 or greater were HER2 IHC 3+ irrespective of the ratio. Seventy-five percent (15 of 20) of the cases with ratio less than 2.0 were HER2 IHC 3+.¹⁰  Interestingly, in that study, IHC 3+ staining was evaluated in 20 of 21 cases and reported as being focal, heterogeneous, or homogeneous in 10 of 20 cases, and only 66.7% were reported to have homogeneous staining, in part explaining the discrepancy with our rates.¹⁰  In our cohort, 81% of targeted HER2 FISH recounts resulted in a HER2-positive result, making them eligible for HER2-targeted therapy.

In this study, 68 of 142 HER2 FISH group 3 cases (47.9%) were classified as HER2 negative after additional workup, including HER2 IHC and targeted blinded FISH recounting in HER2 IHC equivocal/2+ cases, which would have been classified as HER2 amplified/HER2 positive following the 2013 ASCO/CAP guideline.¹¹  Sixteen of the 86 cases (19%) had average HER2 copy number less than 6.0 in the targeted recounting of the IHC 2+ areas despite the original overall average HER2 copy number being 6.0 or greater. There was heterogeneity within tumors in both protein expression, reflected in variability of IHC staining intensity, and amplitude of gene copy numbers in FISH. The assumption that the most intense areas of IHC staining will reliably and reproducibly correspond to the areas of tumor with highest gene copy number may not be accurate or reflect true biology. In addition, analytical variations may be responsible for the observed differences. In cases with 2+ IHC that had targeted blinded recounts, we did not find any association between the intensity (weak versus moderate) and percentage circumferential staining and the likelihood of HER2-positive result (average HER2 signal copy ≥6.0/cell) in targeted areas, suggesting that in the absence of intense circumferential staining (3+), there may not be a good correlation between IHC staining intensity and HER2 gene copy number.

In the decisive HER2 trials, the cases with HER2:CEP17 ratio less than 2.0, including those with average HER2 copy number per cell 6.0 or greater, were considered negative for HER2 amplification and excluded from HER2-targeted therapy. The BCIRG-006 trial included patients if HER2 copy number was 10 or greater and central IHC testing was positive (3+). Originally these cases were considered to have duplication/polysomy of CEP17. However, many of these cases represented HER2 amplification that included regions of centromere rather than chromosome 17 polysomy.^12–18  Although the clinical response of this group to HER2-targeted therapy was limited, the 2013 ASCO/CAP guideline¹¹  recommended that cases with average HER2 copy number 6.0 or greater by single- or dual-probe assay be reported as HER2 positive for gene amplification. The 2018 update⁵  recommended integrated IHC and FISH (in situ hybridization) review for some categories, including results with a positive range (≥2) ratio with low (<4) average HER2 copy number per cell (group 2) and low ratio (<2) with HER2 copy number 6.0 or greater (group 3) or HER2 4 or greater to less than 6 (group 4). The 2018 update moved about half of the cases with HER2:CEP17 ratio of less than 2.0 and average HER2 copy number 6.0 or greater from the group of cancers eligible for HER2-targeted therapy.

In the BCIRG-005 trial, which did not include HER2-targeted therapy, HER2:CEP17 ratio of less than 2.0 and average HER2 copy number 6.0 or greater (ASCO/CAP FISH group 3) cases had similar disease-free and overall survival compared with HER2 FISH–negative (ASCO/CAP FISH group 5) cases.¹⁹  In this study, of the 6 patients who received NAT, only 1 (16.7%) achieved pCR. Four of those 6 patients had final HER2-positive status, and their NAT included HER2-targeted therapy, with 1 of 4 (25%) achieving pCR, lower than the 40% to 70% expected pCR rates for HER2-positive patients and within the range of response expected in HER2-negative patients.^7,20 

A retrospective NAT study from the United Kingdom showed that patients with FISH group 3 tumors that were also IHC 2+ had 12% pCR with HER2-targeted therapy versus 0% pCR when HER2-targeted therapies were not included.⁷  These patients, including the ones who got HER2-targeted treatment, did not have IHC-guided targeted/blinded recounting. The lack of correlation between quantity and intensity of circumferential staining in 2+ IHC and gene amplification in group 3 cases in our study, and the United Kingdom NAT study that offered HER2-targeted therapies to group 3, IHC 2+ patients without a blinded second count achieving pCR in some, suggest that blinded/targeted recounting may be redundant in this subgroup and/or may eliminate rare patients from HER2-targeted therapy eligibility.⁷ 

This study has potential limitations. Most of our cases were from our reference laboratory, with a small number of in-house patients who had clinical follow-up information. However, group 3 results are rare in any one institution; this study would not have been possible without including reference laboratory cases, which by nature have very limited attached clinical information.

To the best of our knowledge this is the largest series of breast cancer cases with ASCO/CAP group 3 HER2 FISH results (HER2:CEP17 ratio of <2.0 and average HER2 copy number ≥6.0). After recommended integrated evaluation of IHC, and, when indicated, targeted blinded FISH recount, 74 of 142 cases (52.1%) had a final HER2-positive status, which is about half of the patients who would have been eligible for targeted therapy following the previous (2013) ASCO/CAP guideline. Average HER2 copy number significantly correlated with final HER2 status. In HER2 IHC equivocal (2+) cases, the extent and intensity of staining did not appear to correlate to the levels of gene amplification, suggesting lack of evidence for IHC guidance in identifying heterogeneously amplified areas in group 3 FISH. In the limited number of patients with clinical follow-up data, HER2-positive status was not associated with a response rate comparable to group 1 (HER2:CEP17 ratio ≥2.0; HER2 copy number ≥4.0) patients. Future prospective studies are needed to determine if some of the additional workup, especially IHC targeted FISH recounting, is helpful in determining HER2 status and identifying patients who will benefit from targeted therapies in this group of patients.