HER-2/neu protein overexpression in breast cancer is mostly caused by HER-2/neu gene amplification. However, it is unclear whether aneusomy 17 may also play a role. Using immunohistochemistry assay (IHC) with DAKO antibody and manual quantitation, 189 specimens were selected from archival invasive breast cancer specimens, including most IHC-positive and some IHC-negative cases (n = 158 and 31, respectively). They were then analyzed by PathVysion fluorescence in situ hybridization (FISH) assay (Vysis, Inc., Downers Grove, IL) and by an image analyzer (ACIS; ChromaVision Medical Systems, Inc., San Juan Capistrano, CA)-assisted IHC quantitation. Ninety-two cases contained disomy 17 (chromosome 17 centromere, 1.76-2.25 signals per cell) whereas 97 cases had aneusomy 17, including 82 with low polysomy (2.26-3.75 signals per cell), 10 with high polysomy (> or =3.76 signals per cell), and 5 with hypodisomy (> or =1.75 signals per cell). HER-2/neu protein expression had the highest correlation with HER-2/neu gene dosage (copy number; r =.826), followed by the HER-2/neu gene to chromosome 17 ratio (r =.733). The lowest correlation was with the chromosome 17 copy number (r =.307), on which the 10 cases with high polysomy 17 had a disproportionately high impact. The FISH assay using the PathVysion criterion for HER-2/neu gene amplification (HER-2/neu gene to chromosome 17 ratio, > or = 2.00) achieved higher concordance with ACIS IHC than did an alternative FISH criterion (absolute HER-2/neu gene copy number, > or = 4.00 signals per cell). Most ACIS IHC-PathVysion FISH-discordant cases contained disomy or low polysomy 17, whereas all 10 cases with high polysomy 17 had no such discordance. However, two cases with monosomy 17 had ACIS IHC-PathVysion FISH discordance, i.e., with gene amplification, but no protein overexpression. Both cases would have had no gene amplification if the alternative FISH criterion had been used. In conclusion, aneusomy 17 is common in breast cancer. Except in a certain subset of cases, aneusomy 17 probably is not a significant factor for HER-2/neu protein expression or for clinical assessment of HER-2/neu status.