Abstract

Aims The current study was done as a validation study prior to setting up a clinical HER2 testing service using the new commercial Poseidon HER2 fluorescence in situ hybridisation (FISH) assay. However, it was felt that the experience of the authors of this study may be of interest to other laboratories when considering setting up a HER2 diagnostic facility.

Methods 122 patients who had been diagnosed with invasive breast cancer were selected. Immunolabelling with HercepTest, PathVysion and Poseidon FISH assays were carried out using tissue microarray blocks.

Results Concordance correlation coefficients showed near perfect agreement in average HER2 and centromere specific signal counts per cell and in the HER2/CEP17 ratios between the PathVysion and the Poseidon FISH assays. In addition, the κ measure showed perfect agreement (κ 0.9441, p<0.0001), and if only 2+ cases were considered there was substantial agreement (κ 0.7671, p=0.0006), between the two assays. The sensitivity and the specificity of the Poseidon FISH kit were calculated to be 95.2% and 100%, respectively, whereas the positive predictive value (PPV) and negative predictive value (NPV) were 100% and 99%, respectively. With regard to the ability to presume HER2 polysomy, the Poseidon FISH kit had a sensitivity of 93.3% and a specificity of 99.1%, with PPV and NPV of 93.3% and 99.1%, respectively, as assessed with PathVysion classification as the reference.

Conclusions Statistical analysis confirmed that the two FISH assays are comparable in terms of detection of HER2 gene amplification. Proceeding from these findings, the genetic diagnoses obtained with the Poseidon kit can be considered to be as valuable as the results from the Food and Drug Administration approved PathVysion assay, and its utilisation in routine HER2 diagnostics is proposed.