PathVysion—a US Food and Drug Administration–approved dual-probe human epidermal growth factor receptor (HER2) fluorescence in situ hybridization (FISH) assay—provides the HER2:CEP17 ratio, a centromeric enumeration probe ratio for determining HER2 status in breast cancers. However, pericentromeric amplifications might then skew the HER2:CEP17 ratio, underestimating the HER2 status, which calls into question the use of CEP17 as the reference probe.
To analyze the utility of a noncentromeric chromosome 17 reference locus (D17S122) to assess HER2 gene status in cases showing “nonclassical” FISH patterns with the CEP17 probe.
The HER2 status of breast cancers accessioned in the years 2015–2017, displaying “nonclassical” or “equivocal” results by the PathVysion (Abbott Molecular Inc, Des Plaines, Illinois) HER2 DNA Probe Kit were reflex tested using an alternate FISH probe (ZytoLight SPEC/D17S122, ZytoVision, Bremerhaven, Germany) and interpreted with American Society of Clinical Oncology/College of American Pathologists 2013 guidelines.
Of 37 cases, 17 were FISH equivocal. With the alternate D17S122 probe, 13 (76.4%) were reclassified as amplified, 3 (17.6%) as nonamplified, and a single case retained an equivocal result. Of the 17 cases with a chromosome 17 polysomy pattern, disomy, polysomy, and monosomy patterns were seen with 14 cases, 2 cases, and 1 case, respectively. Within the 17 cases with polysomy pattern, 3 (17.6%) demonstrated an unusual colocalization pattern of HER2 and CEP17, which was not observed with the alternate probe.
The denominator-stable alternate probe is a useful adjunct in the diagnostic armamentarium to analyze HER2 status in cases with FISH equivocal and complex patterns.
Human epidermal growth factor receptor 2 (HER2) overexpression and/or gene amplification connotes bad prognosis in breast cancer.1 Most importantly, breast cancers with HER2 alterations are treated with targeted anti-HER2 therapies, which have been shown to markedly improve response rate and survival when added to chemotherapy or as monotherapy. The advent of targeted therapy, especially trastuzumab (Herceptin, Genentech, South San Francisco, California) marked a revolutionary change in the management of early and metastatic HER2+ breast cancers.1,2 It also predicts sensitivity to anthracycline-based chemotherapy regimens as well as relative resistance to cyclophosphamide-based regimens and hormonal treatment.1 The potential cardiotoxicity and exorbitant cost of the targeted treatment raise concern. Also, studies have shown that approximately 20% of the HER2 tests presently performed are fallacious.3 Hence, an accurate estimation of HER2 status is critical in all invasive breast cancer cases for appropriate patient management.
Although immunohistochemistry (IHC) and fluorescence in situ hybridization (FISH) are the most widely used testing methods, FISH is more objective for HER2 status determination.4 Many laboratories use IHC as the primary assay for screening HER2 status, followed by reflex FISH testing in cases of IHC HER2 equivocal (score 2+) cases.
The most commonly used FISH assay is the dual-probe assay, 1 fluorescently labeled probe hybridizing to the target HER2 gene, and 1 probe hybridizing to the centromeric region of chromosome 17 on which the HER2 gene is located.5 The latter probe, CEP17, is used as a surrogate marker for the chromosome 17 number, and thus increased CEP17 signals above a certain threshold are considered to indicate the presence of polysomy 17.6 However, recent studies with broader genomic approaches, including comparative genomic hybridization and multiplex ligation–dependent probe amplification, have demonstrated the rarity of true polysomy in breast cancers, and increased CEP17 signals were thought to be due to the pericentromeric amplifications.7 Thus, CEP17 might not be an appropriate reference gene for determining the status of chromosome 17 and can mask true HER2 amplification status.6
Taking this issue into consideration, the current American Society of Clinical Oncology/College of American Pathologists (ASCO/CAP) 2013 guidelines4 have provided diagnostic criteria considering both the ratio of HER2 to CEP17 and the absolute number of HER2 signals. Thus, cases are termed as “amplified” for HER2 gene when the HER2:CEP17 ratio is greater than 2 or mean HER2 copy number is greater than 6, when the HER2:CEP17 ratio is less than 2. These criteria ensure the maximum accuracy of HER2 assessment. However, FISH equivocal cases constitute another diagnostically gray area that needs additional testing to confirm whether the patient is eligible for anti-HER2 treatment. To further investigate equivocal HER2 FISH results (HER2:CEP17 <2 with mean HER2 copies numbering 4 to <6), the ASCO/CAP 2013 guidelines4 proposed employing an alternative probe to CEP17 that is not expected to coamplify with HER2. However, this guideline specified neither the probe nor the diagnostic criteria to be employed while using the alternative probe.
With the FISH technique, most cases are objectively and unequivocally interpreted.8 However, there is a small subset of “unusual/nonclassical” cases (like colocalization of HER2 and CEP17 signals,9 chromosome 17 aneusomies, low level of HER2 gene amplifications, etc) that poses significant diagnostic and interpretive difficulties.8 These cases highlight the underlying mechanisms of genomic alterations and their implications in the determination of HER2 status.8 Previous studies have shown that using an alternative reference probe for chromosome 17 for assessing cases with complex HER2:CEP17 and equivocal FISH patterns has significantly altered the HER2 status.5,6,10,11
With this preamble, we aim to analyze the utility of an alternate, noncentromeric chromosome 17 reference locus to assess HER2 gene status accurately in cases with unusual/nonclassical FISH patterns. Because the D17S122 gene, located on the short arm of chromosome 17 (17p12) is a stable and conserved gene locus,12 we selected HER2/D17S122 as an alternate reference probe for HER2. This approach would be helpful in further stratifying patients accurately as being HER2+ or HER2− in determining whether they should be recipients for anti-HER2 targeted therapy.
MATERIALS AND METHODS
The study was approved by the institutional ethics committee of Tata Memorial Centre, Mumbai, India. The study samples were selected from 1205 patients with primary breast cancers who underwent dual-probe FISH testing for HER2 gene status in the Division of Molecular Pathology, Tata Memorial Centre, from January 1, 2015–January 31, 2017. Results from both diagnostic core biopsies and surgical excision specimens were included.
HER2 FISH testing was performed using the PathVysion (Abbott Molecular Inc, Des Plaines, Illinois) HER2 DNA Probe Kit and the ZytoLight SPEC HER2/CEN17 Dual Color Probe (ZytoVision, Bremerhaven, Germany) according to the manufacturer's protocol. The slides were scored by an expert molecular pathologist using an Olympus BX53F upright fluorescence microscope with a QICAM (Q34130) Olympus camera (Tokyo, Japan) and Q-Capture Pro 7.0 image analyzer software (Rockville, Maryland). At least 40 breast cancer nuclei were scored per case. In cases with FISH equivocal results or complex FISH patterns, like tumor heterogeneity, polysomy patterns, etc, an additional 20 to 40 cancer nuclei were counted. Also, in the latter scenario, the slides were additionally scored for at least 40 cancer nuclei by another expert molecular pathologist. HER2 FISH scoring was interpreted as amplified, equivocal, or nonamplified as per the ASCO/CAP 2013 guidelines.4 The FISH reports of all cases were retrieved, and data on the HER2:CEP17 ratio and the mean HER2 and mean CEP17 copies per cell were recorded for each case. The FISH results were further classified into various categories as per the previous study13 and are as shown in Table 1.
Definitions of Human Epidermal Growth Factor Receptor (HER) Fluorescence In Situ Hybridization (FISH) Groups Used per Ballard et al13 Criteria

Inclusion criteria:
Cases with nonclassical and equivocal FISH results (Table 1);
Cases with classical FISH results but unusual FISH patterns, like chromosome 17 aneusomies (chromosome 17 polysomy pattern was defined as an average of ≥3 CEP17 signals7,14 ; chromosome 17 monosomy pattern was defined as an average of <1.4 CEP17 signals), and with cluster genomic heterogeneity (cases with clusters of amplified cells that comprise at least 10% of the tumorous population).7,13
Cases meeting the inclusion criteria formed the study group. The study samples were further tested using the alternate reference probe ZytoLight SPEC ERBB2/D17S122 Dual Color Probe (ZytoVision) using the manufacturer's instructions. Figure 1 depicts the probe design of both the HER2/CEP17 probes (Figure 1, A) and the HER2/D17S122 probes (Figure 1, B). The sections were scored manually by at least 2 expert molecular pathologists using an Olympus BX53F upright fluorescence microscope with a QICAM (Q34130) Olympus camera and Q-Capture Pro 7.0 image analyzer software. For all cases, at least 40 nonoverlapping tumor nuclei were counted. The FISH results were interpreted in accordance with the ASCO/CAP 2013 guidelines and were classified as follows: HER2/D17S122 greater than 2 was amplified; HER2/D17S122 less than 2 with a mean number of HER2 copies greater than 6 was amplified; HER2/D17S122 less than 2 with a mean of 4 to 6 HER2 copies was equivocal; and HER2/D17S122 less than 2 with a mean number of HER2 copies less than 4 was nonamplified. The cases were further subclassified into different subgroups as per the previous study13 (Table 1).
Probe design for human epidermal growth factor receptor 2 (HER2) gene by fluorescence in situ hybridization test, with (A) centromeric CEP17 locus as the reference gene, and (B) alternate, noncentromeric D17S122 locus as the reference gene.
Probe design for human epidermal growth factor receptor 2 (HER2) gene by fluorescence in situ hybridization test, with (A) centromeric CEP17 locus as the reference gene, and (B) alternate, noncentromeric D17S122 locus as the reference gene.
The results of the HER2 FISH test with both of the probe sets were compared to analyze whether reflex testing with the alternate, noncentromeric D17S122 loci altered the HER2 status. Also, changes in the FISH groups were noted in cases with unaltered HER2 status. Furthermore, for analysis of chromosome 17 aneusomies, comparison of different patterns (monosomy, disomy, and polysomy) with both probe sets was done.
The following clinical and histopathologic details for the patients (wherever available) were obtained from the electronic medical records: age at diagnosis, sex, tumor size and grade, histologic subtype, lymph node and distant metastasis, IHC details of hormone (estrogen receptor and progesterone receptor), and HER2 receptors. The IHC details of the antibodies used are as enumerated in Table 2.
RESULTS
Clinicopathologic Features
The study group comprised 37 cases. All patients were women in the age group of 30 to 74 years, with a median age of 49 years. The tumor size ranged from 1.2 to 10 cm, with a median size of 3.6 cm. All tumors except 1 were invasive breast carcinoma, no special type, grade III. A single case was invasive micropapillary carcinoma, grade III. Of 33 cases, lymph node metastasis was present in 24 cases (72.7%), and distant metastasis was present in 6 of 37 cases (16.2%).
Classification of Cases Using CEP17 as Reference Gene
Of the total 1205 breast cancer cases tested for HER2 gene status using the centromeric HER2/CEP17 reference probe, 37 cases (3%) met the study criteria and were included in the present analysis. Initial interpretation of 37 cases with HER2 FISH testing with the centromeric probe was amplified for 14 cases (37.8%), equivocal for 17 cases (45.9%), and nonamplified for 6 cases (16.2%). These cases were further classified using the criteria by Ballard et al13 as “classical” and “nonclassical” FISH results (Table 1). The 11 cases with nonclassical FISH results were further stratified as coamplified (n = 5; 45.5%), monosomy (n = 4; 36.4%), or low amplified (n = 2; 18.2%). Of 9 cases with classical FISH results, 7 cases (77.8%) had unusual FISH patterns; chromosome 17 polysomy pattern was seen in 4 cases and cluster heterogeneity was noted in 3 cases. Two cases with classical FISH results and typical FISH patterns were included as control cases (Figure 2).
Schematic representation of the distribution of the classical and the nonclassical fluorescence in situ hybridization (FISH) results within the study group. Abbreviation: HER2, human epidermal growth factor receptor 2.
Schematic representation of the distribution of the classical and the nonclassical fluorescence in situ hybridization (FISH) results within the study group. Abbreviation: HER2, human epidermal growth factor receptor 2.
The clinicopathologic features of the cases according to the FISH results are as shown in Table 3. HER2 protein expression by IHC was equivocal (score 2+) in 32 of 36 cases (88.9%) and positive (score 3+) in the remaining 4 cases (11.1%). One case was IHC uninterpretable even on repetition of the test.
Impact of Retesting With an Alternate Reference (D17S122) Probe
Change in the Final HER2 Status
Of all 37 cases, 1 case was uninterpretable in view of extremely weak signals with the noncentromeric probe and was excluded from further analysis. Reflex HER2/D17S122 testing changed the HER2 interpretation in 16 of 36 cases (44.4%), all of which were initially labeled as equivocal using the CEP17 reference gene. Of 17 FISH equivocal cases, only 1 case (5.9%) retained its equivocal status with the D17S122 reference probe, whereas 13 cases (76.5%) were reclassified as amplified (monosomy, n = 6; low amplified, n = 5; and classical amplified, n = 2; Figure 3) and 3 cases (17.6%) were reclassified as nonamplified (Table 4). All of the FISH equivocal cases were equivocal for HER2 protein expression on IHC as well (Table 3).
Fluorescence in situ hybridization analysis in breast cancer cases using (A) centromeric (CEP17) reference probe and (B) alternate, noncentromeric (D17S122) reference probe. “Equivocal” case (A) was redesignated “classical amplified” case (B) with D17S122 loci. Green arrow showing average red (human epidermal growth factor receptor 2 [HER2]) signals, 4 to 5 per cell, with disomic CEP17 (green) signals, whereas HER2 showed classical amplification with multiple green (HER2) signals (blue arrow) when interpreted in relation to D17S122 reference loci (original magnification ×1000).
Figure 4Fluorescence in situ hybridization analysis showing (A) coamplification, with a colocalization pattern of human epidermal growth factor receptor 2 (HER2; red) and CEP17 (green) signals seen as yellow fused signals (blue arrow). B, However, with D17S122 reference loci, the case was classical amplified, with multiple green (HER2) signals and disomic (red) reference signals (yellow arrow), with loss of colocalization pattern suggestive of pericentromeric amplifications rather than true polysomy in breast cancers (original magnification ×1000).
Figure 5Fluorescence in situ hybridization analysis showing concordant results of chromosome 17 monosomy pattern using both probes, CEP17 and DI7S122, as reference loci. A, White arrow showing most tumor cells show single green CEP17 signals. B, Yellow arrow showing a similar monosomy pattern with single red D17S122 signals in most tumor nuclei (original magnification ×1000).
Fluorescence in situ hybridization analysis in breast cancer cases using (A) centromeric (CEP17) reference probe and (B) alternate, noncentromeric (D17S122) reference probe. “Equivocal” case (A) was redesignated “classical amplified” case (B) with D17S122 loci. Green arrow showing average red (human epidermal growth factor receptor 2 [HER2]) signals, 4 to 5 per cell, with disomic CEP17 (green) signals, whereas HER2 showed classical amplification with multiple green (HER2) signals (blue arrow) when interpreted in relation to D17S122 reference loci (original magnification ×1000).
Figure 4Fluorescence in situ hybridization analysis showing (A) coamplification, with a colocalization pattern of human epidermal growth factor receptor 2 (HER2; red) and CEP17 (green) signals seen as yellow fused signals (blue arrow). B, However, with D17S122 reference loci, the case was classical amplified, with multiple green (HER2) signals and disomic (red) reference signals (yellow arrow), with loss of colocalization pattern suggestive of pericentromeric amplifications rather than true polysomy in breast cancers (original magnification ×1000).
Figure 5Fluorescence in situ hybridization analysis showing concordant results of chromosome 17 monosomy pattern using both probes, CEP17 and DI7S122, as reference loci. A, White arrow showing most tumor cells show single green CEP17 signals. B, Yellow arrow showing a similar monosomy pattern with single red D17S122 signals in most tumor nuclei (original magnification ×1000).
Distribution Across Different FISH Groups
Of 20 cases with concordant HER2 FISH status using both of the probes, change in the distribution of FISH subgroups was noticed in 5 coamplified cases (25%): 4 of these cases (80%) turned out to be classical amplified, and 1 (20%) was low amplified (Table 4). Among 7 classical FISH results with unusual FISH patterns, cluster heterogeneity in 2 of 2 cases (100%) with a cluster of 10% or more amplified tumor cells was noted with both of the probe sets (1 case with cluster heterogeneity was uninterpretable with D17S122 probe), whereas 4 of 4 cases (100%) with chromosome 17 polysomy pattern (average CEP17 signals ≥3) revealed a disomy pattern with the D17S122 reference locus.
Analysis of Chromosome 17 Aneusomies
Of all 36 cases, chromosome 17 polysomy pattern was noted in 17 cases (47.2%), with an average number of CEP17 copies ranging from 3.03 to 9.15, and chromosome 17 monosomy pattern was seen in 4 cases (11.1%).
With an alternate probe, chromosome 17 polysomy pattern was retained in only 2 of 17 cases (11.8%), whereas it showed the disomy and monosomy patterns in 14 cases (82.4%) and 1 case (5.9%), respectively (Table 5). In addition, 3 of the 17 cases (17.6%) with a polysomy pattern demonstrated a unique, unusual colocalization pattern of HER2 and CEP17 with the conventional probe that was not observed with the alternate probe; we published research on 2 of these cases previously9 (Figure 4). Cases with chromosome 17 monosomy pattern (4 of 4; 100%) were concordant with both of the probe sets (Figure 5).
DISCUSSION
To the best of our knowledge, this is the first, single-institution–based study from India evaluating the impact of the alternate probe (D17S122) on HER2 gene status in invasive breast cancer cases tested with a conventional centromeric (CEP17) reference probe. In this study, we subjected the HER2 FISH equivocal cases, cases with nonclassical FISH results, and cases with unusual FISH patterns as evaluated by the US Food and Drug Administration–approved HER2/CEP17 dual-color FISH probe, to reflex testing by an alternate, noncentromeric, HER2/D17S122 FISH probe. The use of this alternate probe resulted in refinement of diagnosis in 16 of 17 HER2/CEP17 FISH equivocal cases (94.1%). A total of 13 cases (76.4%) were reassigned to the amplified category, and 3 cases (17.6%) received a diagnosis of classical nonamplified.
In most invasive breast cancer cases, accurate HER2 status can be detected unequivocally using the FISH technique.1 However, there is often a subset of cases that pose diagnostic and interpretative dilemmas.8 One of the reasons for this might be the use of the CEP17 locus as a reference gene, with α-satellite sequences being more prone to copy number aberrations, thus skewing the HER2:CEP17 ratio.5 To deal with the diagnostic dilemmas that come with using a conventional dual HER2 probe, previous studies attempted to use other loci, such as tumor protein p53 (TP53), Smith-Magenis syndrome critical region (SMS), and retinoic acid receptor α (RARA), as controls for the CEP17 loci.6,10 Tse et al6 demonstrated that 58 of 132 nonamplified cases (43.9%) and 13 of 14 equivocal cases (92.9%) were reassigned to the amplified category using these alternate noncentromeric probes (Table 6). Similarly, Troxell et al10 found amplification in 5 of 7 cases (71.4%) after using alternate chromosome 17 SMS and RARA probes in cases with copy number aberrations affecting the centromeric loci.
Studies in the Literature Using the Alternate, Noncentromeric Probe for Human Epidermal Growth Factor Receptor 2 (HER2) Gene Status Evaluation and Impact of This Evaluation on Final HER2 Status

There was a major impact on the final HER2 status with an increase in the number of FISH equivocal cases and/or amplified cases with the recent ASCO/CAP 2013 guidelines because the previous ASCO/CAP 2007 guidelines used only the HER2:CEP17 ratio for determining HER2 status.11,15 Also, for the first time the guidelines recommended the use of an alternate probe for the retesting of FISH equivocal cases.4
The D17S122 gene, located on the short arm of chromosome 17 (17p12), being well conserved and having a stable locus, is being increasingly used for reflex testing of FISH equivocal cases. Two recent studies—one by Shah et al11 and another by Donaldson et al5—on employing this probe have reported HER2 amplification in 212 of 405 (52.3%) and 39 of 73 (53.4%) FISH equivocal cases, respectively, whereas we noticed amplification in 13 of 17 FISH equivocal cases (76.4%; Table 6). Also, conversion from equivocal to nonamplified cases was seen in 3 of 17 cases (17.6%), which is consonant with the previous studies.5,11 Furthermore, with the alternate probe Donaldson et al5 reported 15 of 170 nonamplified cases (8.8%) as being amplified and 5 of 67 amplified cases (7.5%) as being nonamplified. On the contrary, no results for the amplified or nonamplified cases in the present study were discordant with the alternate probe. Table 6 enumerates previous studies in the literature that incorporated alternate chromosome 17 reference probes for the determination of HER2 status.
The use of the HER2/D17S122 probe revealed the disomy pattern of chromosome 17 in most cases (82.4%) given a diagnosis of polysomy pattern with the CEP17 probe. This is concordant with recent studies showing the rarity of true chromosome 17 polysomy in breast cancers.7 Moreover, the rare colocalization and coamplification pattern of HER2 and CEP17 noticed in 3 of the cases was not seen with the alternate probe. This further supports the hypothesis of extension and incorporation of the HER2 amplicon into the pericentromeric region proposed for this unusual feature.8,9,16 Also, the ease of reporting FISH cases with the alternate probe is evident from the fact that of 5 cases of HER2 coamplification, 4 turned out to be classical amplified.
The potential limitations of the present study are the small sample size (n = 37). Second, we used only 1 noncentromeric, alternative locus (D17S122) for comparison with the centromeric probe, because this is a stable locus.12 However, it is well known that chromosome 17 copy number aberrations are common in breast cancer, and even copy number aberrations can affect this stable locus.5,6,17 Hence, to determine true chromosome 17 status it would be prudent to perform testing with multiple other alternative probes, like RARA, SMS, TP53, etc. Third, the study lacks clinical correlation with anti-HER2 therapy in those cases with revised HER2 status, and monosomy and low-amplified HER2 cases.
To summarize, this series is unique in the sense that we subjected cases with nonclassical FISH patterns and chromosome 17 aneusomies, in addition to FISH equivocal cases, to reflex testing. Reclassification of HER2 status was noted in 16 of 36 cases (44.4%) and in 16 of 17 FISH equivocal cases (94.1%). Thus, from the present study and previous studies,5,11 it is evident that the denominator-stable alternate probe is a useful adjunct in the diagnostic armamentarium to analyze HER2 status, especially for FISH equivocal cases and cases with complex FISH patterns. The goal of the alternative probe assay is to better stratify patients as positive or negative in order to determine whether they should receive HER2-targeted therapy. Proper assignment of HER2 status would help better stratify patients as HER2+ or HER2− and ensure that a patient is treated with the appropriate therapy.
The authors would like to thank Sandeep Dhanavade, HSC, MLT; Vinayak Kadam, HSC, MLT; and Sonali Tambe, BSC, DMLT, for the technical assistance with performing the FISH test.
References
Author notes
The authors have no relevant financial interest in the products or companies described in this article.
Competing Interests
Presented in part at the annual meeting of Association of Molecular Pathology; November 10–12, 2016; Charlotte, North Carolina.