HER2 FISH results in breast cancers with increased CEN17 signals using alternative chromosome 17 probes - reclassifying cases in the equivocal category.

AIMS: HER2 testing of invasive breast cancer by in-situ hybridisation guides therapy decisions. Probing HER2 and cen17 simultaneously is supposed to reveal both a potential HER2 gene amplification and polysomy 17. However, a considerable number of breast cancer patients with quasi polysomy 17 are considered "equivocal" which is diagnostically meaningless. Moreover, patients with equivocal / false polysomic tumours are prevented from a potentially beneficial anti-HER2 treatment. Here we evaluated the RAI1, D17S122, and TP53 hybridisation markers to reliably indicate true polysomy and to accurately reclassify equivocal samples as HER2-positive. METHODS AND RESULTS: Samples with (n = 82) and without (n = 52) increased cen17 copy numbers and 78 evidently HER2-amplified specimens were analysed using dual and triple marker hybridisation probes. Selected putative polysomic samples were subjected to array-based comparative genomic hybridisation. We found that 37.8% samples with putative polysomy 17 did not show any gain in RAI1, D17S122, or TP53. Accordingly, aCGH revealed evidence for the presence of HER2/cen17 coamplification rather than for true polysomy in those cases. Reflex testing using alternate 17p markers reclassified up to 56.3% equivocal cases as HER2-positive and the combined assessment of a 17p and 17q locus allows the differentiation of true vs. false polysomy. CONCLUSIONS: An increased cen17 copy number does not necessarily reflect polysomy and reflex testing facilitates the reclassification of "equivocals". Nevertheless, the prognostic and predictive value of a HER2/cen17 coamplification versus HER2 gene amplification alone must still be prospectively evaluated. This article is protected by copyright. All rights reserved. Twelve DNA samples were extracted from formalin-fixed, paraffin-embedded (FFPE) tissue specimens, tissue cores, obtained from primary breast tumors. All samples were co-hybridized on SurePrint G3 Human CGH Microarrays against reference DNA samples obtained from peripheral blood of healthy female donor. To much the DNA fragment size distribution of test and reference samples prior to labeling and aCGH hybridization all reference samples were sheared to and average DNA fragment size of 500 bp. aCGH analysis was conducted to validate results of FISH analaysis that had been previously obtained for the same clinical samples.