Introduction

In the western world, breast cancer (BC) is the most commonly diagnosed malignancy among women, representing about 30 % of all new cancer cases, and after lung cancer the second leading cause of cancer death [1, 2]. The current cancer care guidelines for BC recommend that estrogen receptor (ER), progesterone receptor (PgR), and human epidermal growth factor receptor 2 (HER2) status must be routinely determined in all patients with invasive BC, BC recurrence and BC metastases [3, 4]. These guidelines were published to help improve laboratory performance in the determination of these markers, which provide useful predictive information regarding response to targeted therapy.

HER2, located on the long arm of chromosome 17 (17q12), is amplified and/or overexpressed in about 15 to 20 % of invasive BC. Cases with a HER2-positive status represent a clinically important subset of BC associated with poor outcome but also with a high likelihood of response to HER2-targeted therapy [5–8]. Several studies have shown that anti-HER2 therapy given during and/or after chemotherapy results in a significant improvement in disease-free and overall survival [9–11]. Therefore, HER2 is a helpful marker for therapy decision making in patients with BC and appropriate evaluation of HER2 status ensures that the right patient receives the right treatment [3].

At present, HER2 protein expression is determined in BC samples by immunohistochemistry (IHC) resulting in three possible outcomes: negative (score 0 or 1+), equivocal (score 2+), and positive (score 3+). If the IHC result is equivocal, reflex testing should be performed on the same specimen using an alternative assay, such as in situ hybridization (ISH) [3].

The new 2013 ASCO/CAP (American Society of Clinical Oncology/College of American Pathologists) guideline has updated the definition of HER2-positive status by modifying both IHC and ISH criteria, reducing the thresholds for post-analytical interpretation of positive results in comparison with the previous 2007 ASCO/CAP guideline [3, 12]. In the new guideline, a HER2 score 3+ is defined as the presence of complete and intense membrane staining, in at least 10 % of tumor cells [3]. This represented a return to the IHC criteria originally used in the first-generation clinical trials [13]. A similar approach was used regarding ISH criteria (see below).

In this study, we aim to compare the impact of the change from the 2007 to the 2013 ASCO/CAP guidelines on the result of HER2 amplification test in BC.

Materials and methods

Cases

A series of 916 primary invasive BC cases was retrieved from the archives of Ipatimup Diagnostics, including cases evaluated 1 year before (494 cases from November 2012 to October 2013) and 1 year after (422 cases from December 2013 to November 2014) the publication of the new ASCO/CAP guideline (November 2013). All BC cases (core biopsies and surgical specimens) had been fixed in 10 % formalin, embedded in paraffin, and were referred to our institution (national reference center for HER2 ISH) with an equivocal IHC HER2 score (2+) to perform the HER2 amplification assay with a HER2 Dual ISH DNA Probe with a silver marker (SISH).

Ethics approval and informed consent were not required for this study.

SISH

SISH testing was performed on 3-μm sections of formalin-fixed, paraffin-embedded tissue of all BC cases using dual-hapten, dual-color ISH (DDISH). The dual-probe assay (INFORM HER2 Dual ISH DNA Probe Cocktail Assay; Ventana Medical Systems, Inc., Tucson, Arizona) contains a HER2 locus-specific probe and a control probe specific for the centromere of chromosome 17 (CEP17). The entire procedure was carried out on an automated staining system (Ventana BenchMarkTM XT Staining System) according to the manufacturer’s instructions. Positive and negative controls were used for each staining run.

Evaluation of the results included recording the number of HER2 and CEP17 signals in at least 20 nuclei in two different areas. The samples were classified by pathologists (AP and FS) according to the 2007 and 2013 ISH criteria for HER2 amplification. Corresponding hematoxylin and eosin staining were used for the identification of the invasive component of the tumor.

The 2007 ASCO/CAP guideline defines HER2 amplification as positive at a HER2/CEP17 ratio >2.2, equivocal at a HER2/CEP17 ratio ≤2.2 and ≥1.8, and negative at a HER2/CEP17 ratio <1.8 [12]. The 2013 ASCO/CAP guideline establishes the result of HER2 amplification as positive at a HER2/CEP17 ratio ≥2.0 or a HER2/CEP17 ratio <2.0 and an average HER2 copy number per cell of ≥6.0, equivocal when HER2/CEP17 ratio <2.0 and average of HER2 copy number ≥4.0 and <6.0 signals per cell, and negative when HER2/CEP17 ratio <2.0 and average HER2 copy number of <4.0 signals per cell [3].

Chromosome 17 polysomy was defined as an average of ≥3.0 CEP17 signals per cell [14]. Genomic heterogeneity was also recorded and considered present if a discrete population of tumor cells with HER2 amplification represented at least 10 % of the total tumor cell population [3].

Statistical analysis

Statistical analyses were performed using the Statistical Package for the Social Sciences (SPSS) version 21.0 for Windows. The Pearson’s chi-square (χ 2) test and McNemar test were used for comparison of qualitative variables and the t test for quantitative variables. The level of significance was set at p < 0.05.

Results

The 916 BC cases concerned 97.2 % women and 1.2 % men. The age ranged from 24 to 103 years, with a median age at diagnosis of 59 years.

The distribution of gender, age, HER2/CEP17 ratio, and average HER2 copy number per cell were not statistically different between the pre- and post-new guideline cases (Table 1 and Fig. 1). The only parameters that changed significantly with the new guideline were the average CEP17 copy number per cell and the presence of chromosome 17 polysomy (4.1 to 0.9 %; p = 0.003; Table 1).

Table 1 Differences between the cases before and after the introduction of the 2013 ASCO/CAP guideline
Full size table
Fig. 1
figure 1

Examples of results of HER2 detection by SISH technique (×400). a HER2-positive; b HER2-negative

Full size image

Table 2 and Fig. 2a present the results of HER2 test performed on the pre-new guideline cases (using the ISH criteria from the 2007 ASCO/CAP guideline): 415 cases (84.0 %) HER2-negative, 18 cases (3.6 %) HER2-equivocal, and 61 cases (12.4 %) HER2-positive. Table 2 and Fig. 2b present the results of HER2 test performed on the post-new guideline cases (using the ISH criteria from the 2013 ASCO/CAP guideline): 348 cases (82.5 %) HER2-negative, 3 cases (0.7 %) HER2-equivocal, and 71 cases (16.8 %) HER2-positive. The differences are statistically significant (Table 2—statistical analysis A; p = 0.003). We also observed that 52.1 % of the positive cases (37/71) fulfill both criteria of HER2/CEP17 ratio ≥2.0 and average of HER2 copy number per cell ≥6.0 (Table 3 and Fig. 2b). We furthermore classified the pre- and post-new guideline cases using the 2007 and 2013 ISH criteria and observed a slight but non-significant increase in HER2-positive cases and a similar decrease in HER2-equivocal cases (Table 2—statistical analysis B and C; p = 0.185 and p = 0.261, respectively).

Table 2 Classification of HER2 test according to the 2007 and 2013 ISH criteria
Full size table
Fig. 2
figure 2

Cases before (a) and after (b) the introduction of the 2013 ASCO/CAP guideline

Full size image
Table 3 Classification of the cases after the 2013 ASCO/CAP guideline
Full size table

In the reclassification of the two case series using the 2007 and 2013 ISH criteria, we observed an increase in HER2-positive cases (12.4 to 14.2 % and 15.9 to 16.8 %, respectively) and a decrease in HER2-equivocal cases (3.6 to 1.6 % and 2.4 to 0.7 %, respectively). This was statistically significant in the pre-new guideline cases (Table 2—statistical analysis D; p = 0.011) and near significant in post-new guideline cases (Table 2—statistical analysis E; p = 0.071).

In the pre-new guideline cases, the 2013 ISH criteria reclassified 22 (4.5 %) of the cases, 9 as HER2-positive (from HER2-equivocal), 7 as HER2-negative (from HER2-equivocal), and 6 as HER2-equivocal (from HER2-negative). All HER2-positive cases according to the 2007 guideline remained HER2 positive with the 2013 guideline (Table 4).

Table 4 Classification of the cases before the 2013 ASCO/CAP guideline
Full size table

In the post-new guideline cases, genomic heterogeneity was detected in 0.47 % of the cases (2/422), the proportion of HER2 amplified cells varying between 25 and 40 % of the tumor cell population.

Discussion

Our center (Ipatimup) is one of the reference centers for SISH test of BC in Portugal. In our center, the introduction of the updated ASCO/CAP guideline for HER2 test by SISH resulted in a significant increase of positive cases (12.4 to 16.8 %) and decrease of equivocal cases (3.6 to 0.7 %).

Several studies recently reported an increase of HER2-positive cases evaluated by FISH but also an increase of HER2-equivocal cases with the introduction of 2013 ASCO/CAP guideline [15–19]. However, the study by the group of Garbar et al. had results similar to ours using FISH, with an increase of HER2-positive cases and a slight decrease in HER2-equivocal cases [20]. The explanation for these differences is not clear, but this might be related to the number of cases, pre-analytical conditions, and different ISH platforms. We did not review centrally the IHC performed externally, which might explain the decrease in equivocal cases in comparison with recent literature.

As yet, the published concordance rates between SISH and FISH vary between 92 and 99 %, the majority fulfilling the ASCO/CAP validation requirement of a concordance rate exceeding 95 % (Table 5) [21–30]. However, the requirement in the 2013 ASCO/CAP guideline to determine the average of HER2 copy number (first applied to bright field ISH and now applied to the FISH test) introduces a problem that did not exist previously. Autofluorescence in FISH might result in overestimation of both HER2 and CEP17 signals, resulting in HER2/CEP17 ratios below 2.0 and average of HER2 copy numbers above 4 per cell and an increase of equivocal HER2 results [31, 32]. If an increase of HER2-equivocal cases by FISH and a decrease of HER2-equivocal cases by SISH is confirmed, the concordance rate of these two ISH tests might decrease to under 95 %. This would open up the question which of these techniques provides the most reliable information on HER2 amplification status.

Table 5 Concordance rates of SISH vs FISH according to the 2007 ASCO/CAP guideline
Full size table

For nearly half of the cases studied (52.1 %), both criteria (HER2/CEP17 ratio and average of HER2 copy number per cell) were fulfilled to allow them to be classified as HER2-positive. This is particularly relevant given the fact that half the cases would be excluded from targeted therapy if HER2 amplification would be evaluated using just the HER2 probe (as some methods do).

Classification of the pre-new guideline and post-new guideline case series with the 2007 and 2013 criteria did not result in significant changes in the HER2 test results. This suggests that modifying the threshold in IHC, from 30 to 10 % of cells with moderate staining, had little effect on the HER2 amplification test results. Lee et al. found that cases with equivocal IHC (score 2+) in 10–30 % of the cells had a probability of being amplified of 5–12 % [33]. It is then not surprising that inclusion of these cases does not significantly change the HER2 amplification test results.

In contrast, classification of pre-new guideline and post-new guideline cases with different ISH criteria (2007 and 2013) resulted in significant changes in HER2 amplification test results. Our findings suggest that the 2013 modified ISH criteria had a stronger impact on the test results than the modified IHC criteria. We found that the 2013 ISH criteria resulted in reclassification of 4.5 % of the cases. Other publications have shown a reclassification rate of up to 15 % of cases [16, 17].

Polysomy of chromosome 17 changed from 4.1 to 0.9 % with the introduction of the 2013 ASCO/CAP guideline, which is probably due to modification of the definition of equivocal IHC HER2 staining (score 2+) rather than a change in the biology of the tumors. Several studies have shown that polysomy of chromosome 17 (measured on the basis of CEP17) varies between 3 and 49 % of the cases, depending on the definitions of polysomy and on the method used [12, 14, 34, 35]. The approach is based on the notion that CEP17 copy number is a surrogate marker for chromosome 17 copy number. However, molecular karyotyping has revealed that an increased CEP17 signal number is usually due to gain of the pericentromeric region rather than to duplication of the entire chromosome [36–41]. CEP17 might therefore not be a good marker for polysomy 17, making true polysomy 17 probably a rare event in BC. Nevertheless, CEP17 amplification can still be the cause of misleading HER2 amplification and false-negative test results, excluding patients from anti-HER2 targeted therapy [34].

Tumors with polysomy 17 are thought to be different from non-HER2 amplified tumors, associated with a more aggressive clinical behavior and not responsive to conventional therapy [14, 42]. However, in BC, the relationship between polysomy of chromosome 17 and the response to anti-HER2 therapy remains to be determined [43–45].

We found the presence of genomic heterogeneity to be rare as observed in just 0.47 % of cases. Several studies have addressed this issue in the past and reported genomic heterogeneity in 5 to 40 % of BC cases [14, 46–49]. Studies on the relationship between genomic heterogeneity and prognosis have shown that tumors with a HER2 amplification in at least 30 % of the cells have a reduced disease-free survival [48, 49]. However, the definition of genomic heterogeneity has also changed from individual cells (between 5 and 50 % of tumor cells with HER2 amplification) to discrete populations of tumor cells (at least 10 % of the total tumor cell population with HER2 amplification) [3, 50]. Additional work is needed to determine the prevalence of genomic heterogeneity with this new definition and the response to anti-HER2 targeted therapy in these patients.

In conclusion, we show that the new HER2 guideline results in an increased number of HER2-positive and a decreased number of HER2-equivocal cases using the SISH technique, primarily because of modifications of ISH rather than of IHC criteria. As a consequence, the 2013 ASCO/CAP guideline selects more patients for anti-HER2 targeted therapy.