br Tumors demonstrating co amplification of multiple
Tumors demonstrating co-amplification of multiple 2-NBDG were further examined by simultaneous hybridization using two separate probes (labeled with distinct fluorescent markers) against the genes of interest, permitting assessment of the co-existence of the amplified genes or chromosomal regions in single nuclei or single amplicons. Removal of protein from the tissue sections, denaturation, hybridization, and post-hybridization washing were performed as described previously . The tissue sections were counterstained with DAPI II (Abbott) and examined using a fluorescence microscope (Olympus, Tokyo, Japan) equipped with a Triple Bandpass Filter set (Abbott) for DAPI II, SpectrumOrange™, and SpectrumGreen™, and a filter set specific for SpectrumOrange™ or SpectrumGreen™.
Scoring and evaluation of FISH slides was performed man-ually by counting the target gene signals and control signals in 40 tumor cell nuclei per sample. Gene amplification was deter-mined by modifying the updated guideline for HER2 FISH categories of the American Society of Clinical Oncology/Col-lege of American Pathologists criteria , as follows. Classi-cal amplification status was assigned to specimens that exhibited both a ratio of gene/chromosome ≥2.0 and a ratio of mean gene/cell ≥6.0. These specimens were further classi-fied as HSR type if the amplified signals were clustered, as DM type if the amplified gene signals were scattered, or as mixed (MX) type if both types of signal were observed. ‘Co-amplified/polysomy’ (CoPoly) was assigned to specimens that exhibited a gene/chromosome ratiob 2.0 and a mean gene/cell ratio ≥ 6; ‘monosomy’ (Mono) was assigned to specimens that exhibited respective ratios of ≥2.0 andb 4.0. Specimens that yielded one or more additional copies of genes compared to control signals and were not sorted into the above categories were originally defined as low-level amplification (LA).
MLPA analyses were performed on 322 of the FFPE breast cancer samples. The 7084 gene loci of the 322 tumors were categorized as 241 ‘amplified’, 665 ‘gain’, 6161 ‘normal’, and 17 ‘lost’ genes, as shown in Table 1. Although there was an overall good agreement between MLPA and FISH data, the concordance rate varied among the respective genes, as
shown in Table 1. Concerning ERBB2, there were two cases in which the ERBB2 MLPA results were discordant with those obtained by IHC and FISH: specifically, one ‘LB+’ tumor and one ‘HER2’ tumor showed ‘normal’ ERBB2 values by MLPA. In total, 109 of 322 tumors (33.9%) displayed FISH-confirmed gene amplification of at least one of the 22 examined genes.
Although no significant differences were found between the results of FISH performed on FFPE sections and on imprinted cells, analysis with higher resolution and more pre-cise enumeration was possible in the latter compared to the for-mer, as shown in Fig. 1 (compare A and B, C and D, and G and H, respectively).
Amplification of the genes located on 8p and 8q is summa-rized in Fig. 2A. Amplifications of FGFR1 and ZNF703 were observed at similar frequencies. By dual-color FISH, the sig-nals for ZNF703, FGFR1, ADAM5, and IKBKB, if amplified, were found to be closely associated, suggesting location on the same amplicons. The observed amplicons were of various types, including HSR (Fig. 1A), CoPoly (Fig. 1C), LA (Fig. 1G), and Mono (Fig. 1I). However, except for eight cases of HSR type, all of the amplicons exhibited the PC subtype (Fig. 2A), and thus were considered to be entopically located. In Case 8 (Fig. 1C-F), the tumor nuclei had large signals for centromere 8 surrounded by numerous amplified gene signals for 8p (Fig. 1D, large arrows); this specimen also yielded small paired signals for the 8p genes and centromere 8 (Fig. 1D, small arrows).
Although MYC also was amplified in a variety of amplicon types, DM (Fig. 1J) and CoPoly types were observed the most frequently (Fig. 2A). The PC subtype was detected only in two LA-type and one CoPoly amplicons. In Case 53, tight or loosely clustered signals for MYC and centromere 8 were ob-served (Fig. 1K and L). Co-localization of MYC and MTDH, and of MYC and PRDM14, in single nuclei were found in ten and four cases, respectively. Co-amplification of the genes located on 8p and 8q was observed in seven tumors; however, in each case the two amplicons were separated.
The amplification of CCND1 was found in 40 tumors; 31 of these tumors exhibited the HSR type and 21 of the HSR-type amplicons were of the PC subtype (Fig. 2B). In Cases 69 and 76, increased numbers of CCND1 signals associated with small signals for centromere 11 were observed, as shown
Amplicons in breast cancers