Unraveling the chromosome 17 patterns of FISH in interphase nuclei: an in-depth analysis of the HER2amplicon and chromosome 17 centromere by karyotyping, FISH and M-FISH in breast cancer cells

Chromosome17 (Chr17) is the second most gene-dense chromosome in the human genome [1], containing many genes central to breast cancer development and progression, including oncogenes (HER2, TOP2A, STARD3, TAU), tumor suppressor genes (TP53, BRCA1, HIC-1) and DNA double-strand break repair genes (RDM1) [2‐7]. In particular, the HER2 gene mapping to 17q11-q12 is amplified in 15-20% of all breast cancers [8], it is a prognostic marker for aggressiveness [8] and predicts the response to anti-HER2 agents [8]. An accurate and definitive reporting of HER2 status is thus essential for appropriate treatment determination. Fluorescence in situ hybridization (FISH) with dual probes for HER2 and for the Chr17 centromere (CEP17) is the technique most frequently used in diagnostic pathology to determine the HER2 gene status in interphase nuclei. The correction of HER2 gene copy number using CEP17 signals is required to account for Chr17 polysomy. However, by microarray-based comparative genomic hybridization (CGH) analysis we have recently provided the first direct evidence that true Chr17 polysomy is a rare event in breast cancer [9]. Indeed, a number of CEP17 copies greater than 3 detected by FISH analysis is frequently related to either a gain or amplification of the centromere region, providing another line of evidence that Chr17 usually displays very complex rearrangements.

CGH, loss of heterozygosity (LOH), and molecular genetics studies have shown that Chr17 is rearranged in at least 30% of breast tumors [1, 10, 11] and presents a number of rearrangement breakpoints mapping to either its short or long arm. In particular, 17p is principally involved in losses, whereas CGH on 17q shows complex combinations of overlapping gains and losses [1, 12]. In addition, CGH and spectral karyotyping (SKY) studies have shown that Chr17 is one of the chromosomes most frequently involved in translocations [13]. However the frequency and type of specific structural Chr17 alterations in breast cancer have not been fully detailed. For example, balanced chromosome rearrangements (e.g. translocations or inversions) and low frequency mosaicisms are assessable on metaphases using G-banding karyotype and multicolor fluorescence in situ hybridization (M-FISH) only.

The complexity of Chr17 rearrangements calls into question the accuracy of HER2/CEP17 ratios evaluated on interphase nuclei for diagnostic purposes. Indeed, unsuspected Chr17 rearrangements may be contributing to the equivocal results following in situ hybridization testing, which account for about 10% of all IHC score 2+ carcinomas [14].

By comparing the HER2/CEP17 FISH pattern in metaphase versus interphase nuclei, the present study demonstrated that a CEP17 signal rarely corresponds to a single intact Chr17, in both HER2+ and HER2- cell lines. It is well known that cells of long-term cultures may show high chromosomal rearrangements, however Chr17 was altered even in the short-term TNBC primary culture. Although obtained in a single primary cell line this specific finding may corroborate our hypothesis that chromosomal alterations involving Chr17 in breast cancer may be indeed very complex and merit further investigation in primary cell cultures obtained from carcinomas of different phenotypes.

This extensive cytogenetic analysis of Chr17 demonstrated that only the MDA-MB231 triple negative showed true polysomy (normal chromosome acquisition) as the only Chr17 alteration. The BT474 HER2-positive cells showed Chr17 polysomy together with different Chr17 rearrangements. In ER+/HER2 not amplified cell lines normal copies of Chr17 coexist with rearranged Chr17 copies that either harbor or do not harbor the HER2 gene. On the other hand, some of the HER2 amplified cell lines did not show any normal copies of Chr17 and the HER2-STARD3 gene clusters were observed as HSR on complex rearranged chromosomes. In addition, the Chr17 derivatives (carrying CEP17) did not always show HER2 gene clusters, and these latter were not exclusively observed in Chr17 derivatives. In particular, in the BT474 cells 7 of 9 HER2 gene clusters were found on derivatives lacking CEP17.

In breast cancer specimens the analysis of the HER2 gene in interphase nuclei is requested after an equivocal immunohistochemical result (score 2+) of HER2 expression [8, 19]. Either double-signal (HER2 and CEP17 probes) or single-signal (HER2 probe) assays may be performed assessing the HER2/CEP17 ratio or the absolute HER2 copy number, respectively. In the case of the double-signal assay, the recent ASCO/CAP guidelines [19] recommend using the HER2/CEP17 ratio to screen for amplified or not amplified breast cancers. However, our data provide another line of evidence that the interpretation of the results of the ISH analyses on interphase nuclei using a dual-signal assay should be performed “with caution”, given the high frequency of complex Chr17 abnormalities involving both CEP17 and HER2 loci. This has to be taken particularly into account in cases showing an increased number of discrete HER2 signals, in which the HER2/CEP17 ratio may highly impact on the final definition of the gene status, in contrast with cases with HER2 gene clusters of amplification. The introduction by the ASCO/CAP 2013 of an algorithm that takes into account the ratio first and then the HER2 gene copy numbers represents an improvement in the identification of HER2 positive tumors by dual-signal assays. We should point out that the sole HER2 gene copy number method, as used in single-signal assay, best identifies HER2 gene amplification in interphase nuclei, as CEP17 copy numbers do not reflect Chr17 copy numbers. This finding should be taken particularly into account for those scenarios in which monosomy of Chr17 may be encountered. In this respect the new guidelines seem to be controversial depending on the method of ISH analysis employed. Indeed, by following the single signal copy number method cases with low copy number (<4) are labeled as negative, while the same tumors showing monosomy of Chr17 are labeled as positive if the HER2/CEP17 ratio is employed [19‐21]. Although on one side the HER2/CEP17 ratio may still lead to issues when interpreting Chr17 monosomy and HER2 copy numbers would be more reliable, on the other hand we should also acknowledge that double signal assays with CEP17 counts may still provide informative parameters. Interestingly, a recent study on patients treated with anthracycline-based chemotherapy in the neoadjuvant setting has shown that CEP17 duplication strongly correlated with higher pCR rates [22] than did TOP2A and HER2, in both univariable and multivariable analyses. This shows that alteration of CEP17 copy number detectable in interphase nuclei may still represent a prognostic or predictive indicator, although we cannot decipher the real complexity of the rearrangements this chromosome undergoes to. For instance, Chr17 was frequently translocated with chromosome 8 and 11. These chromosomes have been observed in translocations in many breast cancer cell lines [13, 23] and have also been shown to participate in translocation events in cases of primary breast carcinomas [24]. Cytogenetic analysis of primary cultures would be of invaluable help in understanding whether such alterations recapitulate those of primary tumors. In the primary culture here analyzed, chromosome 8 was involved in one of the translocations.

Another finding of our analyses is the invariable co-amplification of HER2 and STARD3 on the same metaphase chromosomes. Increased co-amplification of HER2 and STARD3 has been described to be correlated with acquired lapatinib resistance [23]. On the other hand, the simultaneous amplification of HER2 and TOP2A was only found in SKBR3 cells, where a pattern of amplification distinct from HER2 was identified in one of the Chr17 derivatives. This observation provides another line of evidence that, despite the genomic proximity of HER2 and TOP2A and the observation that TOP2A amplification seems to be restricted to tumors harboring HER2 amplification, these two genes are likely to pertain to separate amplicons, as previously suggested [24, 25]. One may speculate that secondary rearrangements may intervene to separate the two genes from the primary amplicons.

The results of the traditional karyotyping and of FISH and M-FISH assays on metaphase nuclei reported in this study highlight that complex structural alterations of Chr17 encompassing the HER2 gene and CEP17 are common in breast cancer cell lines. This may reflect the scenario found in breast carcinomas, as this finding was also observed in the primary cell culture raised from a TNBC. Taken together these data indicate that the HER2/CEP17 ratio of interphase nuclei, routinely used to select patients for eligibility for anti-HER2 treatment, should be considered with caution and always coupled with the HER2 gene copy number values in order not to misinterpret HER2 gene amplification, as recently updated in the ASCO/CAP 2013 [19].

Further investigation on primary cell cultures would be of invaluable help to allow functional analysis in cells harboring Chr17 rearrangements with respect of response to distinct therapies.